UNITED STATES DEPARTMENT OF COMMERCE PATENT AND TRADEMARK OFFICE - - - - - - - - - - - - - - - - - - x : PUBLIC HEARING ON TECHNICAL : DOCUMENTATION STRATEGY : : - - - - - - - - - - - - - - - - - - x Suite 912 Commissioner's Conference Room Crystal Park Two 2121 Crystal Drive Arlington, Virginia Friday, May 31, 1996 The hearing in the above-entitled matter, commenced, pursuant to notice, at 9:00 a.m. BEFORE: EDWARD KAZENSKE Deputy Assistant Commissioner for Patents JOHN F. TERAPANE, JR. Administrator for the Search and Information Resources Administration HAROLD SMITH Deputy Administrator for the Search and Information Resources Administration ROBERT STOLL Administrator for the Office of Legislative and International Affairs C O N T E N T S OPENING REMARKS PAGE Edward Kazenske 4 PUBLIC TESTIMONY Jeffrey Forman International Business Machines [testifying on behalf of the Information Retrieval Committee of the American Intellectual Property Law Association] 9 James Y. Loving Client Manager International Business Machines 18 Ronald J. Stern and 32 John Mills 35 [testifying on behalf of the Patent Office Professional Association] Derek Jennings President, Sloane Companies 44 William Mercier Chemical Abstracts Service 51 Orville J. Litzsinger Vice President Alliance for American Innovation 59 Calvin Van Sant Patent Searcher The Whitaker Corporation 68 Glen Wise 71 Kenneith Addison President Oklahoma Inventors Congress 81 Donald W. Banner 86 Kenneth Gural Patentec 98 PUBLIC TESTIMONY (Continued) Eric Vacchio 112 Wilson Plunkett Patent & Trademark Depository Library Association 122 Rodger H. Flagg President Express Search, Inc. 130 Cristopher Flagg Private Searcher 138 CLOSING REMARKS Edward Kazenske 145 - - - P R O C E E D I N G S MR. KAZENSKE: I'd like to get started today. My name is Edward Kazenske, the Deputy Assistant Commissioner for Patents. Joining me today on my left is John Terapane, the Administrator for Search and Information Resources Administration, and Harold Smith, the Deputy Administrator for the Search and Information Resource Administration. On my right is Bob Stoll, the Administrator for the Office of Legislation and International Affairs. This is a hearing to receive public comments on a technical documentation strategy for the Patent and Trademark Office. The public was invited to comment on this issue in a notice published on May 2, 1996 in Volume 61, No. 86, of the Federal Register, pages 19611 through 19612. The notice included six questions which those offering written comments and those testifying at this hearing today were asked to address. In the past, the Patent and Trademark Office has relied heavily on paper files of technical documents, mostly patents, classified by subject matter to carry out its patent granting function. Traditional reclassification activities are bound up by the realities of paper. For every classification assigned to a patent, a patent document must be placed in a file and looked at by the examiners whenever that classification is searched. Flagging all potentially useful information in a patent by the thorough assignment of all classifications using all useful classification schemes would be difficult as every additional classification is another piece of paper in the file. Developing or using alternative categorizations of the same subject matter in a paper-based environment is not feasible. The cost of duplicate physical storage and the impracticality of asking our examiners to search two physical arrangements of the same paper documents would be probably overly burdensome. An automated search environment changes all of this. Multiple classification of a document, alternative classification schemes like the IPC, the EPC, and F-Terms, and other delimiters which would usefully characterize the documents all become only indexes in a computer file. These indexes are easily changed and easily used to yield from an automated search patent documents fewer in number and more pertinent to the examiner's needs than would be possible in the paper file. Providing examiners with mechanisms to enable efficient use of such a variety of retrieval means without adding to their burdens is a requirement which can be met. The progress in the Patent and Trademark Office of automation which will enable this change has been very slow, but the pace has been quickening. Automated searching of the full text of most U.S. patents granted since '71 is now available to all examiners and is being used to varying degrees by most. Retrieval of U.S. patent images for searching or analysis is available to about 25 percent of the examiners now, and it will be available to all hopefully by the end of next year. All examiners now have access to images for patent copy printing purposes. Automation is moving everywhere, but for most examiners so is the paper file. This is the reality. The challenge is to devise a long-term strategy that will maximize our resources being expended to our classification operations to enable a growing focus on the use of state-of- the-art techniques. Our mission must be to enrich and make more effective the systems used by our examiners. With limited resources, prioritizing our current valuable resources and keeping fees to reasonable levels in an environment of a continuously growing number of documents and thereby escalating costs, we must take maximum advantage of the work being done by other organizations, both independently and through increased cooperation, to achieve efficiencies, economies of scale and enhanced capacities. We have received six written comments and 15 requests to appear orally here today. Any persons who wish to speak who have not previously informed us of their desire are encouraged to add their names to the list located on the table at the rear of the room. In order to permit all persons to present testimony including those persons signing up today, each speaker will be limited to ten minutes. Those persons who wish to provide additional comments must submit their comments to the Patent and Trademark Office in writing no later than June 7. The speakers have been listed in the order in which requests were received by the Office. When you present your comments, please give your name and address and whether the comments are your own, those of your firm or company, or whether you represent an organization, or are presenting comments on their behalf. Speakers are requested to limit their remarks to the questions presented in the Notice of Hearing and Request for Comments on a Technical Document Strategy. Thank you. The first person to testify this morning is Mr. Jeffrey Forman, International Business Machines. Jeff. MR. FORMAN: Thank you. I am Jeffrey Forman. My address is IBM Corporation, Suite 605, 1735 Jeff Davis Highway, Arlington, Virginia, and I am representing the American Intellectual Property and Law Association. The American Intellectual Property and Law Association welcomes the initiative of the United States Patent and Trademark Office to solicit input on its technical documentation strategy and appreciates the opportunity to submit this testimony. The American Intellectual Property and Law Association is a professional association comprised of more than 10,000 attorneys specializing in the practice of intellectual property law in private practice, in corporate practice, in government service, and in the academic community. AIPLA represents a wide and diverse spectrum of individuals, companies and institutions involved directly or indirectly in the practice of patent, trademark, copyright, unfair competition law as well as other fields affecting intellectual property. The AIPLA has been working with the USPTO for many years to help the USPTO achieve the benefits of automation and gain from adding new search tools to the examiners' repertoire, and also enhancing the overall effectiveness of the U.S. patent system by making these available to the public. We are very pleased with those aspects of the USPTO's plan set forth in your May 2 Federal Register Notice that states that the USPTO will continue to work with other patent offices and will try various private sector tools and systems to improve its documentation, classification, and search programs. While we recognize the difficulty the USPTO faces in allocating its limited employee resources to cope with an ever increasing workload, we were disappointed that last October before having effective substitutes in place, the USPTO decided to stop adding non-U.S. patent documents to the examiners' search files and to reduce the resources for maintaining and improving the U.S. patent classification system. When these planned actions became known last year, the AIPLA took the position that these steps should not be taken until alternatives have proven themselves to be effective substitutes. The resolution passed unanimously by the board of directors on July 22, 1995 stated, and I quote, "Whereas the USPTO has announced an intention to reduce the classification staff by nearly half and whereas computerized tools have not yet proven to perform as well as human concept-based classification, resolved that the AIPLA favors continuing to use persons knowledgeable in the art area to regularly classify and reclassify U.S. and foreign patent publications, to maintain or reduce the current average subclass size, and to reclassify art areas when any subclass becomes too hard for practical, routine, manual searching or no longer accurately reflects the current state of the art area. "Further resolved that the AIPLA favors in principle cooperation between the USPTO and other patent offices in classification and classification studies." A similar resolution was approved in June 1995 by the Intellectual Property and Law Section of the American Bar Association. That resolution states, quote: "Resolved, that the Section of Intellectual Property Law opposes in principle any reduction or termination or redirection of the existing patent classification program in favor of an all electronic reclassification process without first having in place a proven all electronic reclassification process." A good search is the foundation for quality examination, and effective search files are the key to quality searching. The significance of this was highlighted in the USPTO Customer Satisfaction Survey conducted last year by the Bureau of the Census under the federal government's National Performance Review. In the survey section dealing with patent services, the USPTO standard that was rated as the most important by USPTO customers was quote "conduct a thorough search during the patent examination process of relevant U.S. patent, foreign patent literature and non-patent literature contained in PTO search files, and where appropriate a reasonable search of other non-patent literature." The importance value attached to a thorough search was 4.6 on a scale of five where four is very important and five is extremely important. The January 1995 report by Mathis and Associates entitled "Automated Patent System: Is It Meeting Examiners' Needs?" showed that examiners continue to place their greatest reliance on manual searching of patent documents in the shoes. The survey found that 81 percent of examiners searched the shoes on virtually every application they examined as compared to 44 percent for APS. We commend the USPTO for its initiatives such as the trial of the European Patent Office's EPOQUE II search system and those demonstrated last week at the USPTO reengineering briefing to improve and increase the use of computer searching. We believe that these hold great promise for the future. However, it seems inevitable that for a number of years to come, patent searching will be largely paper-based and a viable and up-to-date U.S. patent classification system remain the key to effective searching. No other classification system is as detailed in its subdivisions as the U.S. patent classification system. No other classification is as thorough in its review of older patent documents when classifications are updated and includes placement of these documents in the newly established classifications to the extent that the U.S. patent classification does. No other classification system is as rigorous as the U.S. patent classification system in following principles of classification in placing documents within the system. No other classification system provides a detailed definition for each subdivision to facilitate the understanding of its content. In addition, an effective U.S. Patent Classification System is also a key to being able to do reliable and cost effective text searching on- line. Inclusion of U.S. classification search terms and search strategies establishes the appropriate context for a key word search with a higher degree of reliability and consistency and less time and effort than trying to do so with additional key word search terms. The USPTO's Classified Search and Image Retrieval search system, CSIR, is also dependent on an effective U.S. Patent Classification System. Among the benefits of CSIR is the ability to browse the patents in a set of classifications without having to spend time viewing the same patent more than once. This is done by combining those classifications in an or'd together set. Another CSIR capability that would be diminished without an effective and up-to-date U.S. patent classification system is the ability to combine classifications representing multiple concepts being searched by adding those classifications together and then browsing those patents retrieved that should contain disclosure of the multiple concepts. Accordingly, while the AIPLA fully supports the USPTO in its continuing efforts to seek out new search and documentation tools to improve examiner productivity and achieve higher quality search and examination, we would urge the USPTO to revisit its decision to cut back on patent classification and reclassification resources and its decision to terminate placing foreign patent documents in examiner search files. The USPTO should not remove the tried and true hand tools from the searching toolbox before the new power tools have been fully developed and have proven themselves safe and effective for use. We recognize that continuing the reclassification program and full placement of U.S. and foreign patent documents in the search files will require the continued allocation of scarce resources as compared to the alternatives planned. However, we believe that maintenance of an effective paper search file should be one of the USPTO's highest priorities until the alternatives are proven effective. We welcome the opportunity to continue to work with the USPTO staff in both developing these alternatives and in strengthening the present search systems and tools in as cost effective a manner as possible. Thank you. MR. KAZENSKE: Mr. Forman, I just have one question. Knowing the AIPLA so well and the high priority that they are placing on this and knowing the other initiatives and priorities of the office, how would they view moving those other priorities in view of this such as a pendency issue or automation fronts or other issues facing the office with the limited resources we have? MR. FORMAN: It all has to be considered as a whole, and the loss in quality of examination as a result of not having as effective a search file as we should for the examiners to use today represents a very high cost. MR. KAZENSKE: Knowing your involvement in classification and that, what is your comparison of the U.S. system to the European ECLA system for this recovery of subject matter? MR. FORMAN: I've never searched in ECLA as such. I have searched in the U.S. Patent Classification System, and as I said, there is no classification system including ECLA that goes to the level of detail and does as thorough a job in placing documents fully retrospectively in current classifications as the U.S. Patent Classification System does. MR. KAZENSKE: Thank you. The next individual is a Mr. James Y. Loving, client manager, International Business Machines. MR. LOVING: Good morning. My name is Jim Loving. I'm the client manager for IBM's Government Industry with business responsibilities for the U.S. Department of Commerce. I'm here today to represent the IBM Corporation. My address is 6610 Rockledge Drive, Bethesda, Maryland 20817. Mr. Deputy Commissioner and other distinguished guests, I am pleased to have the opportunity to present to you today on an important topic to the PTO and to IBM. I will speak to you today about exciting developments in technology that make possible the storage, access, analysis and retrieval of very large amounts of data in massive digital libraries. These new technologies could be deployed to ease the task of classifying and retrieving prior art for PTO, its customers, and stakeholders. IBM is an international company headquartered in Armonk, New York, with 1995 annual revenue of $71.9 billion derived from manufacturing of hardware and component technology, software development and licensing, services and maintenance. IBM's intellectual property is a core corporate asset. Abraham Lincoln said the patent system added the fuel of interest to the fire of genius. And this system through the effectiveness of its implementation rewards creative contributions to society. IBM spends over $4 billion annually in research and development. Seeking patent protection on selected inventions while publishing others is part of a vital multipurpose business strategy. The primary purpose of this strategy is to free IBM professionals to perform research and development that leads to the most advanced and most demanded products that can be manufactured and marketed with minimal patent concerns. Some of our research and inventions result in products and services more quickly, speeding the development of commercial products for use in industry. Partly as a result of our large R&D investment, IBM has been awarded the largest number of patents by the PTO for the last three years consecutively. As one of your largest customers and as a leading information technology supplier and consultant to many of your other customers and stakeholders, we share a common interest in seeking solutions to help you, our collective customers, and ourselves solve the problem of more effectively and efficiently executing the patent examination process. We have been working on several projects that directly relate to enhancing the tools available to inventors needing to search prior art. I believe this same technology could be deployed within the Patent Office "factories" to assist patent examiners in this process as well. IBM believes that we are beginning a new phase or era of computing--network computing. Network computing is a powerful change that has sweeping implications for individuals and institutions alike. From a technology perspective, new means of communication and high speed digital links are enabling real time collaboration and universal connectivity allowing businesses and institutions to achieve greater speed to market, greater flexibility and nimbleness, greater global expansion, and greater customer and supplier integration. One of the most profound implications of network computing may be the development of digital libraries around the world making accumulated knowledge of the ages available instantly and universally. Many organizations from cultural institutions like the Vatican Library, media organizations like the Institute for Scientific Information, institutions of higher education like Indiana University, and government institutions like the Library of Congress, to name just a few, are reaping the benefits of new developments in digital libraries that make texts, images, audio and video, available to more people, more often than ever before. Recognizing this imperative, the U.S. government has invested $24.4 million into the advancement of the digital library initiative. The goal of the initiative is to advance the means to collect, store, and organize information in digital forms and make it available for searching, retrieval and processing via communication networks. IBM launched our digital library initiative in March of 1995. It consists of an array of products and services aimed at helping customers transform information into digital multimedia form that can be shared with users via high-speed networks including the Internet. The initiative also gives customers the tools to manage, present, and protect information. The IBM digital library provides electronic equivalence of a traditional library structure. This structure is categorized as content, creation, and capture, storage and management, search and query techniques for information access, information distribution and presentation and rights management. One of our digital library collaborations includes work between IBM and Derwent, the leading specialist in scientific patent information for over 40 years. We have helped them to develop a new, more flexible way of accessing U.S. patent data. As part of this collaboration, IBM developed a scaleable U.S. patent library server. It consists of all issued U.S. patents for the last 20 years and their associated images. From a Web browser or Lotus Notes client, IBM employees and Derwent clients may search this library allowing for text searching of the bibliographic text and claims data with complete viewing of the patent image, full patent image. The data is officially accessed with enhanced searching capabilities such as attribute and recursive queries because it has been indexed with a sophisticated relational data base, DB2. This capability is available to anyone by purchasing a turnkey PC-server and is accessible over the Internet. We would be glad to work with the PTO to provide this capability in support of your automation efforts. The Institute for Scientific Information has been a worldwide leader in providing scientific reference information to the research community. After extensive dialogue with its customers, ISI realized that its users were increasingly asking for more complete information to be delivered directly to their desktops. Given ISI's long-standing reputation within the research community, it was important for the company to meet the needs and demands of its customers who were requesting a solution that could deliver full images and text directly whether it be at home, in an office, or in the laboratory. ISI needed a system that provided navigation tools to access the worldwide scientific literature while ensuring protection of copyrighted contents. IBM and ISI partnered to develop an electronic digital library that offered users direct access to Current Contents Life Science providing table of contents, bibliographic data, and abstracts for over 1350 journals. In addition, users also received full images of articles where publishers have agreed to allow electronic storage and delivery for those journals to which they subscribe. Many organizations that have invested heavily in information technology and stored business information electronically have not always been able to fully capitalize on its value. Advances in the field known as data mining are helping organizations leverage data more effectively and to obtain insightful information. Data mining technology is distinct from executive information systems or decision support systems. These systems typically use a query model to discover facts about data. Data mining uses a different model for the creation of information about the data. It uses a discovery model. Data mining technology helps customers discover and identify previously hidden associations of data and sequential patterns of behavior hidden in large data bases including textual data. One of these specific technologies has been deployed by IBM's European Center for Advanced Mathematics with clients that need to understand relationships among patent data. This technology is called Technology Watch. With this technology, data bases of documents such as patents, scientific publications and technical reports are subjected to automatic analysis and themes are identified and groupings are established identifying documents by themes. Today, Technology Watch is being applied to patent data to discover trends in technology development and invention by answering questions that are important to maintain competitive advantage. This capability of segmentation and clustering could be applied to discover the structure and patterns in very large data sets and used to assist in the classification of new patents and reclassification of patent documents that have been captured and stored digitally. With the move to network computing and the explosion of data that is available for users on the network, it is necessary for users to be able to use search assistance in their quest to find relevant information. This is, of course, at the heart of the problem when considering the patent examination process and the need to find and analyze prior art. IBM has announced several new intelligent information network services applications that assist users in their need to gather pertinent data for large diverse amounts of content that is resident across the network. The InforMarket and InforSage search services are exciting new offerings that will enable organizations that have greater needs to access more data than ever before. Users will connect, search, retrieve, analyze, create and share relevant information in ways that they have previously been unable to do. They are comprehensive and cost effective solutions to information retrieval and provide a simple and consistent way to access any available information source regardless of its geographic separation or retrieval mechanism. With these services, intelligent agent technology and powerful topic profile editors are used to personalize searches across disparate content data bases, and the results are delivered automatically to the users. These search techniques do not require upgrading legacy content data bases and use open architecture and client/server approaches. The user interface can be customized for each client's individual needs. In addition, IBM has been spearheading intellectual property protection for the information that is transmitted through the network through the use and commercial availability of cryptolope containers technology. Cryptolopes are secure packaging for digital information, enabling Internet users to buy and sell contents securely over the network. The innovative application of information technology is playing a key role in the transformation of government. In fact, information technology is in the process of completely reshaping how vital government services are delivered electronically to citizens around the world whether it involves electronic tax filing, renewing driver's licenses or simply requesting government information. Technology improvements in network communication and heightened exposure and acceptance by citizens of on-line services through the Internet place governments in a powerful new position to its constituents. The technology I have been discussing today exists, is commercially available, and is being used to solve customer problems and add value to their organizations. Given the state of the art of this technology, it is possible to digitize and place on-line all significant non-patent literature and patent literature and make them available to patent examiners as well as other customers and stakeholders over the Internet. USPTO search files along with other technical literature could be linked with the search files of different patent offices from around the world. When considering the notion of a patent literature digital library or patent information bank of world patent literature, the significance of such a library would be to enable both inventors and examiners alike to more effectively determine, protect and create intellectual property. Network computing when combined with the ability to establish large digital libraries makes possible the realization of the PTO vision where patent and trademark data is quote "created once, managed effectively, used often, and are accessible over the global information infrastructure." This vision when realized would put the intellectual property of inventors around the world, both published literature, prior art, and all patents and published patent applications, at the fingertips of PTO's employees, customers, and stakeholders everywhere. This is important to PTO and relevant when considering an appropriate strategy for technical documentation and the possibilities for implementing automated information systems that fuel the fire of inventive genius. Thank you. MR. KAZENSKE: Thank you. Questions? MR. TERAPANE: Yes. Mr. Loving, two questions. One is for your indexing with a relational data base, how labor intensive is that, number one? Number two, with regard to use that sort of mirrors the PTO use, and this may be perhaps with your legal department, your intellectual property department, could you give us some more information about the use of this particular data base of U.S. patents by that part of your organization? MR. LOVING: Yes. This was a collaboration that began about a year and a half ago with Derwent. They shared a common problem that we did, which is that we wanted to have more effective access. So we worked in collaboration with the group from our digital library group in California, and there were information technology specialists, data base specialists involved in actually taking the patent data which was sourced from Derwent, which they had bought from the PTO, and loading this data in. So it was a manual process. You'd have to get the data into the data base and index it. That's the first part of the question. The second part is that today this is internally being used by inventors within our research division across a firewall protected Internet to our patent library server, and there are discussions today about making this a public service as well. MR. KAZENSKE: Thank you, Mr. Loving. AUDIENCE PARTICIPANT: Can the floor ask a question? MR. KAZENSKE: No, they may not. Our next speaker--we have a joint speaker here, Mr. Ronald Stern and Mr. John Mills testifying on behalf of Patent Office Professional Association. MR. STERN: Thanks for giving us the opportunity to appear before you today. My name is Ronald Stern. I'm president of the Patent Office Professional Association. We represent all the patent examiners and classifiers at the U.S. Patent and Trademark Office. With me is John Mills, who is a member of our executive committee and is an expert on patent classification. As a labor union, we represent not only the economic interests but also the professional interests of both examiners and patent classifiers. It is in the professional interest of both examiners and classifiers to have a retrieval system that is both efficient and provides 100 percent recall of the relevant art. No one likes to look at irrelevant documents and no one wants to be embarrassed by missing a relevant reference. Right now the U.S. Patent Classification System provides the greatest efficiency and the most relevant documents more often than any other system we have had contact with. Consequently, the Association strongly recommends continuing and maintaining the United States Patent Classification System classifying foreign and non-patent literature documents into that system until, as the AIPLA wisely recognized, until there is a tested and proven alternative. I just want to make two quick comments, and the rest of the testimony will be given by John. I want to highlight one particular area. With respect to replacement of the United States Patent Classification System by ECLA, I wonder if consideration has been given to the extra search time that would be necessary. Today, a U.S. patent examiner reads and understands and searches a case in about four to five hours. A European patent examiner, a searcher, does the same job in about 17 hours. If we are going to balance the cost of the--well, let me just rephrase that. When balanced against the cost of the extra search time required for using the European system, the cost of maintaining our classification system seems dirt cheap. I just cannot imagine the Patent Office telling examiners they will have four times as much time as they currently have. That cost would be astronomical. There is one other item I want to say quickly, and that is that the failure to classify and place foreign patents and other kinds of non-patent literature into our search files is a matter of extreme concern to most examiners. We cannot do a quality job without ready access to those documents. Let me turn over the podium to John, and he'll go into the technical details. MR. MILLS: I'd like to focus briefly on the issues asked and presented in the Federal Register in the notice for this hearing, and for each of these issues, I'm going to read the issue and give our employees' response as we've determined by consensus. And additionally, any remarks that I have made today will be further elaborated on in our written testimony in response to these points. With regard to the first question which the Register calls question A, I'll read it. Are the classification efforts of others using systems or data bases of classification different from the United States Patent Classification System able to replace or partially replace the U.S. Patent Classification System? And we believe--our membership believes--that the answer to this is affirmatively no. No other classification system that we know of, and that would include ECLA, the IPC, and the Japanese patent classification systems, is equivalent or comparable to the U.S. Patent Classification System, and therefore cannot replace it. And substantiating this, we were referred to a document that was prepared comparing those four systems by a Mr. Gary Auton, who works in our division, and, in fact, has a detailed report comparing these systems, the various divergences of them. And that would be included in our remarks, I might add. Additionally, just as a segue with Ron's comments, there are quite a few studies regarding comparisons of use of these systems to the U.S. system, and again I'll reiterate that the European ECLA examiners average about 17 hours per case per search, while our average for our U.S. examiners is about 4.4 hours per case per search. Partial replacement--that's another alternative being considered--is also unworkable proposition. What's required here is the comprehensive classification system, as we already have, and then obviously it has a detailed set of rules and regulations for placement of documents which can be reproducible according to DUPACS scheme, which will be detailed also in our written comments. The idea, of course, of the system is reproducibility whereby two people could search the data base and come up with the same search results. This is quite commonly achievable in the U.S. system but not in the other systems. Often it is true that the U.S. patent classifiers do borrow some of the schedule of ECLA when it's found to be useful, but again we develop definitions and use the rules, the patent classification rules, for placement to firm up and make for reproducibility of patent searching and classification. Question B, or two: Are public users of the classified technical information systems of the Patent and Trademark Office willing to contribute or contribute more toward the ever increasing cost of maintaining these systems? We have no data to the contrary that would indicate that they would not pay for services that were valuable and meritorious. We can't presume that they would, but so far based on our understandings and conversations with users, they seem to be happy with the services provided by the classification division. This includes CD-ROMs, APS text and image searching, as well as the U.S. Patent Classification System searched through the search room. Question 3: Since the usefulness of automated search tools is affected by the characteristics of the technology being searched, should the Patent and Trademark Office tailor search systems by technology rather than maintaining the USPCS for all technologies? We believe the answer is clearly no. The PTO should continue to maintain a single comprehensive data base that provides all users the confidence that a search is complete when it's conducted in accordance with the rules of DUPACS and guidelines of the USPCS. The alternative, balkanizing search files into different data bases with different rules, different procedures, different hierarchies, would lead to a hodgepodge and confusion in conducting a search. Question 4: Since it is known that different search tools and techniques lead to different search results, different results from text or structure search versus classified image search, are automated assisted searches acceptable if there is an overall improvement in the quality of the search? And here this question would appear to be an obvious yes if there is an improvement, but the thing that has to be weighed is the cost of the improvement. Clearly, it's not been shown that any of these tools will result in an improvement. There have been no tests conducted in accordance with information retrieval standards such as standards of precision and recall common in information retrieval art and science. Until the Patent Office has conducted such tests, it's hard to say or impossible to say whether there would be an improvement. We contend that before any replacement alternative is adopted, that such tests be conducted and thoroughly analyzed. And here again I refer to the remarks of Mr. Forman and the quote from the ABA as well as the AIPLA that until we abandon the tried and true tools, we should not elect the power tool. Question 5: Advances contemplated in future search systems permit the examiner to add value to the data base while conducting a search by adding or changing classifications of a document. That is by inserting markers and notes on a document, adding documents, et cetera. This value added capability leads to search files that exist only in electronic form and which are constantly being upgraded. This question has two issues. The first is does the public need access to the examiner's search file? Under this issue, we heartily agree that, yes, the public is always entitled to know what the examiner searched, and this is another argument for the USPCS system. Finally, I'll move to the last question--I see that I have one minute to go--How should the Patent and Trademark Office approach the issue of classification and retrieval of prior art in view of the current state of the art in electronic document search and retrieval area? This question poses the problem that text searching is somehow going to surpass or replace the concept searching of the Patent and Trademark Office Classification System. Our system is designed to protect concepts and ideas. That's what the patent system is all about. Our search system, also our classification system, is designed around concepts and ideas, not around expressions, of the ideas in text format. Text searching is found not to be as accurate as concept and idea searching, and this concept program is done through the definitions of the classification system. No other system has a series of definitions to elaborate on what the meanings are of each subclass. And so on that, I'd just like to affirm that our system is the one that we should follow and maintain. MR. STERN: We are available for questions if you want to ask questions. MR. KAZENSKE: I have one question, and I'll leave it to the rest of the panel. Ron, you made a statement that under this, our examiners cite the most relevant documents in the citation of art. Where is that basis from in your opening? MR. STERN: No, no, no. What--maybe I didn't express myself clearly. What I said was that in terms of a comparison of search systems that when we use the United States Patent Classification Systems, that has been the most efficient way to get the most relevant documents. MR. KAZENSKE: My follow-up to that is since Mr. Will is so intimately involved with our classification system, your opinion of -- MR. MILLS: Mr. Mills. MR. KAZENSKE: Mr. Wills. MR. STERN: Mills. M-I-L-L-S. MR. KAZENSKE: Oh, excuse me--Mills--excuse me. Over the past 15 years or so, how do you view the classification of the foreign documents into that system and the retrieval of that by the examiners over a long period of time here? MR. MILLS: Well, prior to about a year ago, foreign documents were incorporated into the examiner's search shoes and files and then were searched along with in the USPCS system in the convenient and proper way just as U.S. documents were. MR. KAZENSKE: And you believe the integrity of that is complete? MR. MILLS: Well, the integrity may not have been as complete as it might have been if all documents had been placed there. But the integrity now is completely lacking because the documents are not being placed there at all. MR. STERN: That's the essence of it is something is better than nothing, and we'd rather have at least some of the patents there in the right place than none of the patents there. MR. KAZENSKE: Thank you. MR. MILLS: And I might add that the public searchers are also concerned about and the corporations that they represent are concerned about the loss of this prior art. MR. KAZENSKE: Thank you very much. Our next speaker is Mr. Derek Jennings. MR. JENNINGS: Good morning, Assistant Commissioner, distinguished members of the morning panel, and members of the public. My name is Derek Sloane Jennings. I'm the president of the Sloane Companies who has two divisions. The primary division is the patent prosecution services division located at 454 M Street, S.W., Suite 1000, Washington, D.C. Telephone number 202/321-1238. The Sloane Companies conduct prior art searches for all disciplines and prepares prior art reports. I am also a former primary examiner with the PTO from 1982 through 1991. I stand here today as a member of the public who is skilled in providing a valuable service to my clients and who has knowledge to address the important issues and present comments on such issues as outlined in the commissioner's request for comments on technical documentation strategy. The first question I am asked, are the technical documentation classification efforts of others able to replace either in whole or in part the current USPCS and should the PTO use search systems based on technology rather than the USPCS? The PTO has the world's best classification system. It allows for adding or changing the classification of issued patents. It has the means to properly classify technology by the use and the process. All of us know that certain inventions can be used in several fields of technology. The current USPCS has the means for making sure that if a person in one field is considering the question of patentability, that he or she will make sure that all fields are covered before filing for a patent application. Furthermore, since many inventions fall into several fields of technology, the PTO needs the use of the USPCS to make sure that these inventions can be found when the issues of patentability, validity and infringement are raised. Another question I am asked: Is the public willing to contribute or contribute more for using the classified technical information system? Well, this begs two points that I wasn't sure what the commissioner was trying to achieve. If the commissioner's intention is to suggest that a fee for using the classified U.S. patents as maintained in the public search room, such a fee would violate 35 USC 41, paragraph (I)(1), which reads: the commissioner shall maintain for use by the public paper or microfilm collections of U.S. patents, foreign patents, documents, and United States trademark registrations arranged to permit search and retrieval of information. The commissioner may not impose fees directly for the use of such collections or for the use of the public patent or trademark search rooms. Furthermore, if the commissioner's intended suggestion was to increase the fees to cover for the extended costs of maintaining the current USPCS, then I respectfully submit that the PTO is a fee-supported agency, that nominal fee increases for using such systems as APS and imaging work stations at the Public Search and Image Retrieval Facilities would be acceptable if the PTO can justify any increase due to the fact that the PTO fee income from fiscal year ending 1995 was about $577,699,818, as quoted by the Internet Patent News Service. Another question I am asked: Are automated system searches acceptable? Regarding automated system searches, if such techniques are used in combination with a full patent examiner's prior art search, then the automated assistant searches will improve the quality of the patentability search. However, if the examiner solely relies on the automated search, then the issue of patentability quality is called into question because no automated search system can be relied upon unless the automated search system uses a thinking machine with a learning neural network. Should examiners have the capability of adding or changing information on a data base quote "while conducting a search"? Having an examiner perform classification actions while conducting a search would put increased pressures on the examiner who is otherwise under increased pressures to produce a quality product while at the same time increasing the quantity of the work product. Using my own firsthand experience, I have experienced the production pressures and expectations that were required of me during the time that I was an examiner at the primary level. Taking such pressures and expectations, then combining them with outside pressures, I made the mistake of inappropriately taking actions to meet the necessary production goals and subsequently had to leave the office. To place any additional workload on the examiner without relieving of the required production goals cannot lead to a value added capability. In conclusion, the issue of prior art retrieval is an art form. The retrieval of prior art is primarily obtained based on the classification of the technology using the current USPCS. The locating of prior art is best done by noting the location of the classes and subclasses, and to maintain the classes and subclasses, the PTO must retain a full searchable paper collection in order to address all issues of patentability, validity, and infringement. I wish to note that the following comments will be also presented in written form along with comments of others that I've collected from the Internet Patent News Group, and I would thank you for allowing me the time to address these issues. Any questions? MR. KAZENSKE: I have one question. You made a statement that the U.S. Classification System is the best. What is your experience in working in that system and for how long, and following that, what is your experience in working in numerous other systems around the world and how long? MR. JENNINGS: Having been an examiner for almost nine years, I have worked in classes 307, which anybody knows is a collection of electrical arts that covers many areas of electrical technologies. Having looked at those many areas of technologies, I have a very intimate knowledge of class, subclass locations for just about any electrical art form or some mechanical. I feel that it's the best because it allows for multiple reproductions of searches and maintaining or I should say or retrieving a similar search result. I've used Internet service to conduct classification searches, so to speak, out of the UPO and have attempted to do such out of the JPO, and I have not been successful in obtaining a quality search product compared to the USPCS system. So that is what I would say my experience is based on why I think it's better, yeah, in a real world application. MR. KAZENSKE: Any other questions? Thank you, Mr. Jenning. The next speaker is Mr. William Mercier. MR. MERCIER: Good morning. My name is William Mercier, 12620 Fantasia Drive, Herndon, Virginia. I'm speaking on behalf of Chemical Abstract Service, 2540 Olentangy River Road in Columbus, Ohio. Chemical Abstract Services is a not-for-profit scientific data base producer. We are a division of the American Chemical Society, a non- profit scientific membership organization. CAS employs 1300 people in Columbus as well as marketing and sales representatives worldwide. CAS produces the most comprehensive data bases of chemical information and has been known worldwide as the publisher of chemical abstracts since 1907. Information referenced in CA and associated data bases is the result of expert document analysis and indexing by specially trained scientists. My remarks this morning in response to the Federal Register Notice are in regard to the availability of technical documentation from CIS. I would like to address on-line data base offerings, their classification, search tools and their usefulness, cost to management and items that we see as critical for the PTO to consider when contemplating increased reliance on the use of on-line data bases and less reliance on the foreign patents in the USPCS. STN International is a service provided by Chemical Abstract Service in a partnership with Fachinformationszentrum Karlsruhe in Germany and the Japanese Information Center for Science and Technology in Tokyo. STN is an on-line system of approximately 200 scientific data bases covering all areas of technology. STN is extensively used by all segments of the scientific community worldwide. It is also used quite extensively by the PTO's STIC, the electronic information center, the scientific library, and by many examiners in the patent examining corps. A few highlights of the 200 data bases are as follows: the CAS family of data bases covering chemistry, biochemistry, and chemical engineering sources including chemical patents from 20 patent issuing authorities, more than 8,000 journals and conference proceedings, other serials and monographs, technical reports, dissertations and books. The CA Plus File contains over 12 million abstracts and the CAS registry file contains some 15 million chemical substance records. Also, on STN, the Derwent World Patent's Index covering seven million patents from 40 countries. The data base contains abstracts and use advantage statements and for many of the patent citation representative images and drawings. INPADOC, the International Patent Documentation Center data base, covering patent families from 60 countries. The data base contains over 22 million patents. Other patent data bases on STN containing the abstracts and claims of European patents, German patents, and PCT applications. Comprehensive literature data bases for prior art searching such as INSPEC from the Institute of Electrical Engineers, Medline from the National Library of Medicine, Biosis containing biological abstracts and biological reports, the Computerized Engineering Index and the Science Citation Index. All these data bases are searchable using Messenger Command Language. Messenger is a powerful search and display protocol with great power and agility. It features all levels of proximity searching, multi-data base searching, duplicate removal and answer set merging, data extraction, and analysis tools, saving of search queries and customized search profiles. It has expert and novice operating moods with automatic help messages in addition to key word searching. Messenger features powerful chemical structure and substructure, Markush searching, protein and nucleic acid sequence and sub-sequence searching, and automatic thesaurus vocabulary relationship searching. CAS has added in Messenger an on-line thesaurus for U.S. classes and for IPCs including hierarchies, text of the codes, and catch phrases. CAS offers art-specific training on STN to PTO examiners. We recognize the need for tailored search systems and customization of search tools specific to examiner's art. The STN training program at USPTO is needs based, and we endeavor to gear the training to the particular needs of the given art unit or group of examiners. All examiners are taught how to find data bases specific to their art and the classification systems associated with each of the most pertinent data bases. In addition to STN International, CAS offers point and click graphically user interface to chemical information via SciFinder. SciFinder requires no command input and therefore no training and provides a task-based approach to information retrieval with simple phrase input, natural language processing, intelligent query formulation, using customized lexicons and data base smarts. Although we do not yet have a graphically user interface specifically designed for patent examiners, we have presented to PTO mock-ups of task-based interactions designed for them. CAS also offers easy World Wide Web interfaces to patents and very soon an expanded easy World Wide Web interface to STN International. In addition to these off-the-shelf products, CAS offers interfaces designed for individual customers providing integration of commercial data base information with in-house data bases. In order to help manage costs of commercial data base use, CAS has offered USPTO cost management options such as fixed fee plans, site license for communication software and special pricing arrangements for training. We believe that these cost measures are essential for a healthy partnership between CAS and all our major customers. We see three things critical to PTO in its contemplation on the increased reliance on the use of commercial data bases. One, increased awareness for examiners on the breadth of information available and its form. Two, training and support; and, three, employment of modern means of delivery of full text. Examiners must be made aware of the vast array of technical documentation beyond APS. Each group director and SPE must take steps to expose examiners to additional tools for finding prior art including information specialists in the various branches of the STIC, on-line commercial data bases accessible from the desktop, and CD-ROM. Data bases and search systems are valuable only to the extent that they are properly used. Therefore, CAS' emphasis on responsive search support is essential. The quality of help available from the CAS staff of scientists is well-known and highly regarded by users of STN. Until such time as data base systems are truly intelligent, those examiners wishing to do their own searching on commercial data bases must be adequately trained in the system or systems best suited to his or her art. The training should be art specific and should include follow-up and in-group support. Until such time as all scientific data bases are full text, PTO must aggressively pursue modern means for full text delivery of technical documents to the examiner's desk. Increased awareness of the availability of commercial data bases, customized training and support, and modern document delivery will give more the examining corps a means to improve the overall quality of patentability searching. Thank you. MR. KAZENSKE: Thank you. Any questions? MR. TERAPANE: Yes. Bill, as you know, and I know, the chemical area is probably the area where there is the greatest use of on-line searching, and one of the questions deals with whether we should maintain the USPC in the same way across all technologies or consider advances and how that might impact what we're doing? Could you comment further on the use of your data bases or similar data bases in lieu of USPC for doing the type of work that the USPTO does? MR. MERCIER: I can make a general statement about that, but I think the question would have to be more pointed, and that is what kind of art are we talking about? I think in my area, being an organic chemist and working so closely with chemical examiners over the years, I would have to say that indexing by Chemical Abstract Service could and does in many cases adequately replace the USPCS for certain art. Going beyond chemical art, I think that there are very important tools that are sometimes overlooked by examiners and by searchers. The USPCS is one tool, but any professional searcher worth his or her salt will tell you that there are other tools that will find other art and sometimes better art. MR. TERAPANE: Thank you. MR. KAZENSKE: Thank you. Our next speaker is Mr. Orville Litzsinger. MR. LITZSINGER: Good morning, gentlemen. As the vice president of the Alliance for American Innovation, I'm pleased to offer comments on such a vital tool within our patent system for the American patent examiners and their important decision-making process of granting patents. The Alliance for American Innovation represents 10,000 small business and independent inventors from coast to coast and venture capitalists and many companies including biotech companies. Our office is located in Washington, D.C., at 1100 Connecticut Avenue, zip 20036. We need to at least keep the excellent U.S. classification system we now have developed until something that is better comes along to fit the American system. We all know the classification system is one of the primary tools for the examiners in their use of the performance of their professional duties. Since we lead the world in intellectual property, we must be careful not to arbitrarily give away that lead by again giving away and giving up some of our best tools the examiners have to ensure that the quality product is retained. We often think of the Patent Office as the largest technical library in the world. And what good is a good library without a good classification system? I've had significant personal experience with two other critical classification systems in my first life, not patent systems but in analogous situations similar to this, when I was a program director for the United States intercontinental ballistic missiles, and in that system, we had two classification systems for maintaining the technical data and technical information for which to ensure that those systems were kept properly maintained and engineered the way they needed to be. Those missiles sat in silos around the country, and in one case, the technical system that was based to support those missiles and their nuclear warheads was very similar to the system that we have had in the United States up to now, that the examiners used in the Patent Office. It was centrally controlled, it was very detailed and meticulous, it was very thorough. We had high confidence of the output of that system. The silos that housed those missiles by contrast used another system, this time more similar to the system that we might consider going to, a European system or a system which is fragmented and what have you. Therefore, the response to your questions that have been asked are based upon some personal experiences which I have had as well as, of course, the inputs from all of our affiliates, and I will at this time go down those questions and briefly answer each one. On Question A, the answer is that we do not believe the American Patent Office personnel should be put in a position of negotiating with other countries in developing the American system. Receiving good inputs and taking good ideas is fine, but to be arbitrarily placed in a position of having arms twisted in order to harmonize in this area is something that we believe not to be proper. Question B, on paying for maintaining the system, the answer is that American inventors have borne the costs, although sometimes reluctantly, for many fancy ideas which have come from the Patent Office. They have paid for all the employees, your new ideas, only to see potentially the destruction of our classification system and search files. They will object to changing the system and have objected to the ideas of changing the system, and only unless the system changes can give us an upgraded high confidence and thorough system, which U.S. inventors and innovators and our economic sector deserves. Thirdly, Question C, the automated search system has not been satisfactory on all counts. The American system has worked but to go just with word search is not satisfactory. It's inexact plus it creates as many questions as it does answers. Question D, the changes proposed will not, according to the inventors, improve the system. They will cloud it so it is more difficult to find something many times. Question E, the patent examiner search files are private, but the public files are accessible and so are the examiner's files with permission. There is no need, according to the inventors, to change this procedure. Question F, the present system has been successful. New technology for technology's sake does not answer the need for a classification system. We need to remember the purpose of the office is to issue patents and not be chasing the latest technology prematurely. The recommended changes further serve to weaken our system if they continue to mimic foreign systems which are not as thorough as ours. Put the foreign paper back in the shoes now. We endorse the comments of the POPA and others that have and will testify to ensure that the U.S. system is not weakened. Thank you for the opportunity to comment. MR. KAZENSKE: Thank you. Questions? I have a question. You made the comment that we should keep this system until we have a better system. Do you believe there's a role of the USPTO to work at and provide resources for providing a better system? MR. LITZSINGER: Of course it is. The PTO has been operating in the black, and we've been giving our money that we have paid into the PTO to other agencies in the United States government from which they have taken from our source, and there seems to be resources available to improve our system. MR. KAZENSKE: So we should put our resources to that? MR. LITZSINGER: Well, I'm not going to ask you a question. Yes. MR. KAZENSKE: In view of that, you said we have this best system. Should, therefore, the customers that pay into the system to develop that better system be the only ones that contribute to the cost of that future development then? MR. LITZSINGER: Well, that question is a little bit different twist on one of the questions that was asked here. Again, the fees that are already paid are paid by the inventors and the users of the system primarily, and that system supports that issue right now. MR. KAZENSKE: So other parts of the world should be able to use this without an equal contribution to a future system? MR. LITZSINGER: Well, we're so anxious to harmonize. When we talk about fees, maybe that's the area we should be harmonizing first with when we talk about people in other parts of the world. Right now it costs about $5,000 according to the GE study to maintain a patent in the United States, and that cost is about $32,000 in Japan and it's about $132,000 in Europe. So if we start talking about harmonization of costs, I think we'd have to address it from a broader area than just the area that we're approaching. MR. STOLL: I have a question. The prior speaker just indicated that possibly using CAS on-line, an area such as organic chemistry may be better searchable through those methods. Is it possible that some areas are already ripe for electronic services in lieu of the American patent classification system? MR. LITZSINGER: I think all areas are ripe for electronic services if it can serve us in a better fashion than what we have now. But it certainly cannot be implemented without adequate testing for thoroughness as well as operability. So, you know, I don't think that we should segment one area and say, hey, this area is ripe for that. However, if you field test something or case test it, you might choose an area that's most adaptable to doing that. But I would hate to suggest or think that we would consider using different systems for different areas. MR. KAZENSKE: I have one follow-up question. If you had the authority with the resources and to make priority, would you maintain the current system or develop a future system? MR. LITZSINGER: That's a faulty dilemma. Okay. MR. KAZENSKE: What would you do? MR. LITZSINGER: I think that you can't answer that question one way or the other because that's not the real world. You cannot say I'm going to stop the current system so that I will develop a new one. What you do is you would phase in a new system at the same time taking time to ensure that you test it properly and that you have the same confidence level in that new system being as thorough as what you have in the present system, and thus far there is no evidence that we have that. MR. KAZENSKE: So you would, therefore, not put all the resources to the current system? MR. LITZSINGER: We have excess resources to what is being spent on the current system. I said that before. I would like to say that again in case it was missed. We're operating in the black in the Patent Office. We're taking Patent Office money that's being paid in by inventors and we're financing other bureaucracies in this town. So it is not a decision that has to be made that way. MR. KAZENSKE: Thank you very much. MR. LITZSINGER: You're welcome. MR. KAZENSKE: Our next speaker is Mr. Calvin Van Sant. MR. VAN SANT: Good morning. I'm Calvin Van Sant from the Whitaker Corporation, 4515 New Linden Hill Road in Wilmington, Delaware. Although I represent the Whitaker Corporation, which is the patent arm of Amp, Incorporated, an electronics manufacturer, my views are basically just from a public searcher's point of view on how I use the USPCS and such. The one question I do want to address briefly is the question on any improvement the examiners would make to an electronic file, if that would be available to the public. Currently, when I go through the paper shoes, there are several notes here and there on the pages written in pencil by the examiners, which I find very useful. I think in a perfect world if I was searching, I'd like the CSIR. If you could take the CSIR system, make it more efficient to search, make it available to the public with the examiner's remarks, foreign documents, anything that's available, you know, I believe the examiner should be skilled in the art, they should be aware of the art, and they should put that in appropriate places with whether it's the U.S. Classification System or if you can draw up something better, but take those tools in a CSIR format and make it available to the public. Currently, I'm heavy user of CSIR when I'm in the PTO, and the $50 an hour to me is reasonable because it's an efficient search. I can go across several subclasses. I can do key word. I can combine with subclass, but taking away that subclass code, I'm lost. Just key words alone aren't enough. But I do believe whatever you implement and make available to the public has to be an image-based system. Key word searching I miss too much. I rely upon the images. And as far as making the foreign art available, recent searches on validity of U.S. patents I've come up with good art from foreign countries that is published that is missed by the PTO. So I think the foreign art, any art available, is valuable, in a nutshell. MR. KAZENSKE: Questions? I have one question. Have you searched any other classification systems of art? MR. VAN SANT: I am just now in my office trying to develop a system where I can search European and Japanese art on compact disks. I have recreated the paper files of the USPTO for my major subclasses, classes in my office. The IPC works. Like I said, I use it in combination. I can't just trust the U.S. I got to go U.S., European and Japan, and I have limited experience with the IPC. In some areas, it's better, some areas it's worse. That's my only real other experience with the IPC. MR. KAZENSKE: Thank you. The next speaker is Mr. Glen Wise. MR. WISE: Good morning, Commissioner Kazenske, ladies and gentlemen. I am a self-employed patent agent. I've been practicing primarily with a search practice for over 35 years. Before that I was an examiner and a classifier in the Patent Office. I also am an inventor, and I hold over a dozen U.S. patents. Today I will appear on behalf of not only myself but in behalf of many professional searchers and others who utilize the PTO's public search facilities who have made their opinions known to me and asked me if I'd pass them along at this meeting. In addition, in the course of my travels around the PTO, many members of the examining corps and other PTO staff members have made their opinions known to me on subjects being discussed here today and others that haven't been discussed yet, and I will refer to those opinions in my remarks. I forgot to mention my address is 6450 Georgetown Pike in McLean, Virginia 22101, 703/356-1819. First, I would like to offer my comments on the questions as proposed in the notice in the Federal Register and the OG, and incidentally I might remark that that OG notice, which is the main one depended upon by the vast number of practitioners, went out awfully late for response. Some of them--most of them--are probably sitting in the mailrooms or in distribution systems around the country, and I hope they get to the people in time to respond by the June 7 deadline. Beginning with the questions--I'm not going to repeat the questions because they've been repeated several times, but I will paraphrase. Should we give up the USPC in lieu of something else? My answer to that is a definite no also to go along with most of the other speakers if not all the other speakers, but I would say that there are areas where it can be supplemented by some of the other systems. As an example, in Class 424, the medicine area, medical composition, I happen to know that some of the areas are much more detailed than in the U.S. Patent Classification System in the European system. So if we supplement our system with pieces of the other system, we might be much better served and take the best from theirs and add it to ours which is the best. You ought to have no doubt of that. I've used, tried to use the IPC on numerous, primarily on validity searches, and found it wanting in almost every case. It is either not the art you're looking for in the shoes or there is some drawback. For example, it's spread across seven or eight IPC areas and to search all of them, you'd be searching a mountain of art, and you'd have to look it up individually by number because it wouldn't all be in the USPC schedule, of course, in a single subclass. We also have to remember here, and nobody has mentioned to my knowledge yet at this meeting, that the USPC serves a community beyond the Patent Office, and that is the PTDLs. The PTDLs draw in the hundreds of thousands of visitors every year, whereas the public search area draws only in the tens of thousands. So let us just say we gave up the USPC. We're going to ask people who have visited the PTDLs and are somewhat familiar with the system to on their own try to figure out another system, and the last I knew, for example, in the IPC, you had to look in six editions of manuals, just to be sure you'd covered everything, and the IPC also does not backfile when they reclassify. In other words, they look, they start at date Jan 1 '96 say and go forward, but they don't reclassify into their previous subclasses everything that was done before. You have to look in different manuals to find those things. In addition to the comments I'm making, yesterday I filed nine different documents with a cover letter, and I'll refer to that as I go along. Two of the documents relate to drafts of papers written by Bill Lawson, predecessor to Mr. Terapane, just before he left the Patent Office. One is called "Future Search," and the other is called "USPC Versus IPC," and Bill in both cases gives a fantastically good presentation, as far as I'm concerned, of the advantages and disadvantages of both systems and comes down hard on the side of the USPC in both cases. Also, Patent Office newsletters of various types such as one entitled "With Our Classification" in the Patent Office Professional Association Newsletter of January and February of '95 is a good expose of where we stand versus where other classification systems stand. So, just in summary, again, on Question A, I would say keep the present system intact with its evolution as has been going on but supplement it where you can find these little better areas in other systems. Question B: Are the public users of the classified technical information system of the Patent Office willing to contribute or contribute more? I read this one because it should be noted that we already contribute substantially in paying for photocopies, APS services, U.S. patent copies, foreign patent access, and there's a proposal, as I understand it, to charge debit card systems on the CSIS system and on the assignment record system. I don't know where that stands at the moment because I thought it was going to already be installed, but apparently something is holding it up. Also, getting back to the PTDLs, they at this moment and the people they serve pay very little into the system. There are 84 Patent and Trademark Depository Libraries, and I said before there are hundreds of thousands of people moving in there. That might be a source of revenue that has been overlooked all these years, although they're going to complain bitterly, I'm sure, if they get charged in a public library just like we would here if we were charged to walk in the door. Question C, which has to do with the usefulness of automated systems, I would say possibly that Question C should be looked at but should not be implemented without a great deal of study. Now certainly they're going to--in the future there will be areas that are found to be computer searchable only where you get pretty good results, but I think that's in the future somewhat, and until that time and until it's proven, the USPC should be maintained in full force. I found Question D to be a bad question because if you substitute like the word "monkey" for "technical search systems," you'd still come out with a yes, you know, and so I'm just going to skip Question D. Question E, adding value to references and the ability to switch them back and forth has been something that's been done, as one gentlemen mentioned, in paper right along. We value those little notes on the side that say "nylon, see line 36" or whatever. And the implementation of that into the computer system should be urged, and, in fact, it's on the wish list of the subcommittee of the AIPLA. I do some work with as a high priority item. Part one of that, does the public need access to examiner's search files? They certainly do until integrity checks might be run of both the search and the public file because the files are different, as I pointed out to many of the members of the staff of the PTO. There are many more unofficial cross-references--well, let's put it this way-- what used to be called unofficial cross-references in the examiner's files. They've now been officialized. In many cases, you'll find many more patents in the examiner's files than in the search room files. And for post-prosecution litigation purposes, does the public need to know what is in the examiner search file at the time the examiner made the search? Yes, they do. This often comes up in litigation where the one side is trying to prove that the examiner should have seen the reference, and the other side is trying to prove that he shouldn't or couldn't because it wasn't there or whatever. And the office has kept microfilm back for years to show that on a given day, a given patent was in a given place or should have been in a given place. Now with the electronic capabilities, you can definitely say that patent is in that data base on that day, and unless the system was down or something, there is no reason why you missed it, and in that regard, I might remark that as far as I know the Patent Office has been and is still keeping every key stroke that the examiners do. Is that still true; do you know? MR. KAZENSKE: I do not know. MR. WISE: You do not know. Okay. I'll ask the chief information officer next time I see him. I didn't get a chance before this. Talking about--this is a typical search page out of the file, and that hasn't changed since Noah built the Ark, I don't believe, but it should because many examiners now tell me that they just don't have time to delve into the foreign, for example, or they're only doing APS, and some of the questionnaires that I read that have gone around within the office suggest that some examiners, each examiner has his own way of doing business within his own area. I would suggest that this search sheet should be modified to show whether he searched U.S., he or she searched the U.S., and the U.S. or foreign. That's the way we usually write our report--searched U.S. and foreign art and searched U.S. art only, searched U.S. patents--I see my time is up. And I would just remark that--two more things-- under 35 USC 41, as mentioned by Mr. Forman, it appears that foreign documents should be put back into the shoes as well as under Section 44 of the United States Code 3506(d), Paperwork Reduction Act, would seem to require that, as well as your dissemination policy which in paragraph three, as appeared in December 1 '92 Official Gazette, suggests that we should get those documents back unless you want to have a search system with no art to look at eventually, and try to keep quality up that way. MR. KAZENSKE: Thank you very much. MR. WISE: Any questions? Thank you. MR. KAZENSKE: The next speaker is Mr. Kenneith Addison. MR. ADDISON: Congratulations to you, sir, you did not trip over my name; the spelling is correct. It is an old Scotch spelling of the word Kenneth. I go by the name of Ken. I'm the president of the Oklahoma Inventors Congress, the secretary and the director of the United Inventors Association of the USA. We represent roughly 110,000 inventors from throughout the nation. It's a pleasure to be here today and I want to thank you and commend you for the allowance of ten minutes. Having spoken before a number of governmental committees and agencies before, I am accustomed to tailoring my presentation so that I can cut off 25, 50 or 75 percent in order to accommodate the time that is allowed. I believe that in this instance I shall return you a bit of change. Mr. Commissioner, Mr. Deputy Commissioner Kazenske, and distinguished members of the panel, ladies and gentlemen, good morning. It gives me a great deal of pleasure to be here this morning representing all these inventors from throughout this vast nation on a topic which we believe to be of equally vast importance, that of the maintenance of our United States system of patent classifications. We've been requested to respond to a number of questions which were prepared by the PTO relative to the desirability of maintaining this system as the operational key to the efforts of the Patent and Trademark Office. Our concern is not only based on the problems which would be created by the deletion of this system as it relates to the inventive community so much as it is to the efforts of the examining corps to provide efficient, accurate, pertinent and legal responses to patent applicants. Therefore, our responses to your questions, which we will respond to in the order in which they appeared only as Question A, B, C and D, are as follows. Question A: we have always and will continue to borrow ideas from other countries, but those other ideas must meet the higher American standard which has been enlarged to meet the ever growing needs of American technological development. As frequent users of the system, we do not believe that examiners should be placed in a position of negotiating with other countries to develop the American system which is already superior to theirs. Question B: while it has caused a definite hardship in the inventive community, inventors have quote- unquote "gone along and not rocked the boat" because they essentially have an innate fear of objecting to a Patent Office policy. They will, however, strenuously object to changing the classification system and being charged for it. Inventors do believe that patent examiners need a constantly upgraded classification system because technology is growing. But we do not need to throw away the baby with the bath water by destroying the classification system. Question C: Mr. Deputy Commissioner, I can answer this question in the same manner which I addressed patent harmonization when I appeared before USPTO hearings in 1993, and I quote, "If it ain't broke, don't fix it." Question D: There is no system nor is there likely to be a system which will replace the decision-making capability of the human mind. Artificial intelligence would only cloud the issue and the ultimate decision would still rest with the human mind. What have we gained for our dollar? Question E: There is a great deal of justifiable concern in the creation of a private examiner system. One of the fears being that the additions and improvements to the system may somehow not find their way into the public search files while this is particularly onerous when it is considered that the creative public will be paying for those files. The procedure of asking permission to use the examiner's files has worked well and should continue to do so. We also question the wisdom of instituting a system where all the users cannot be equally served by the system or where special electronic access or equipment may be required. Question F: Key word searches do not do it in patents. Close only counts in horseshoes and hand grenades. The present system has been successful since the early 1900s. We need and must keep it. Mr. Deputy Commissioner, adjacent to this building sits the world's largest and most comprehensive technological and scientific library. It is a growing and vibrant thing. We must accommodate that growth, but it must be done with growth in the traditional system which we have all come to know, to trust and to love. To do away with the U.S. classification system would be tantamount to taking a book with 5,500,000 pages, none of which were numbered, which contained no index, and no table of contents, and then closing that book without the benefit of a bookmark. Thank you. MR. KAZENSKE: Thank you very much. MR. ADDISON: Questions? MR. KAZENSKE: No questions. Thank you very much, Mr. Addison. MR. ADDISON: Thank you, gentlemen. MR. KAZENSKE: The next speaker is Mr. Donald Banner. MR. BANNER: Good morning, gentlemen. MR. KAZENSKE: Good morning. MR. BANNER: Friends, Ken Addison is a hard act to follow, but my name is Donald Banner. My residence is 2734 Unicorn Lane, N.W., in Washington, D.C. I represent no organization. I express my own views and the views of several other attorneys around the country who primarily are engaged in litigation. It's been a very interesting morning not only from what I've heard but also based on my experience as the head of a large multinational corporation for many years. One of the things I haven't heard very much about is the reference to the dependability of the system and dependability of the search and dependability of the report is absolutely fundamental. We can't forget. We have in the statute a presumption of validity. That presumption of validity is based on what you people do. Presumption can be ruined over the years if you don't have a reliable, dependable source. That's important to industry. You can't have a situation in which this system of validity is clouded. You know you can't build half a factory. You either know what you're doing or you don't know what you're doing. Indeed, the question of infringement is essential. How do I find out whether or not a certain product I have in mind is an infringement of a patent? I have to go to a proper data base. I have to know whether the infringement is actually of a valid patent, whether the patent is invalid. That requires, I think, a claim based search system. It requires a definite, clear definition of classes and subclasses so that I can know where I can go. It's not a question of whether we're going to cheapen the system. It's not a question of whether it is more expensive to do it the way we're doing it. I noticed in the background statement given to us with respect to this hearing today, really the only suggestion that change might be required is based on limited financial and personnel resources. We can cheapen the system if we want to. It would be devastating, I think, if we cheapen the system. You can't do that and have a dependable system. We could cheapen the system by letting every tenth patent issue without examination. But we all know that would be ridiculous. I find in the background statement also reference to the fact that we are no longer classifying new foreign patents. I think that's a real problem. I think it may very well be considered the inverse of the breach of the duty of candor. You mean if the Patent Office knows about material prior art that the public doesn't know about, that it is concealing from the public by failing to put it into the proper place in the Patent Office, where would the public search? The public would search in the proper area as designated in the manual. How are they going to find these recently issuing foreign patents that are relevant that should be there but aren't? I think that's a grave problem, and I think it's a breach of professional responsibility. Going to the questions, I'm not going to read them, but I will comment upon certain parts of it. The first question talks about partially replacing U.S. Patent System. How do you partially replace a classification system? I don't think that makes any sense. You know either it's--how do you determine whether somebody is partially pregnant? You know either it's there or it's not there. I think you can supplement certain areas if you want to, but you have to be able to have a system in the final analysis throughout the Patent Office where a potential infringer can go and look at the claims that are there and find out whether he's an infringer and have to go to those claims to find out whether that patent is valid. Now, I think it was Mr. Wise that mentioned the fact that he's used European systems in an attempt to invalidate, to see if a patent is invalid, and he said that's a difficulty. I want to ask you a question. Have you ever tried to go the foreign systems and do an infringement search? I have. Try that sometime. When somebody comes to you and says I'm going to build this factory and make all these gizmos, and it's going to cost me $87 million, am I going to infringe anybody's patent or not? Now, business executives can take bad news. Business executives can take good news. But they can't take "gee, I wish I knew." They look at you and say why am I paying you all this money? What is it? That's ridiculous. We can't do that. In the United States, we have a reasonably good system of finding out whether or not we're going to infringe. Try that in foreign patent systems. You can make a decision, of course, but I suspect that after you give the information to your client, he'll go to church and pray that you're right because it's a very difficult thing to do. I don't think you need to replace the present system. I think that's really not feasible, certainly the way we do it, the classification systems of other countries today. Maybe ten years from now they'll have some better way to do it, but they don't today. The second question about increasing costs. Will the public pay more money for that? You know that's a very interesting question to me. In 1996, an applicant who files a patent application pays costs to the Patent Office. In order to obtain and maintain that patent through its normal life, he will pay twice as much as the applicant did in 1990. Now don't tell me that the Patent Office is going broke. That doesn't make any sense. Would the public be willing to pay increased costs? Well, I think the question of that is much more fundamental. What are you doing with the other money? Is the Patent Office being run efficiently? Are there things that the Patent Office is spending their money on that are a waste of time which make patent examiners happy or maybe patent officials happy? I know they're all very happy--and I want them to be happy. But the question of whether you have to increase costs anyplace is a question of what the efficiency of your whole system is. I think if you had to make a judgment, we certainly want to have a dependable, reliable patent system, not only in issuance of patents but also in determining what they cover and what they don't cover. Should they be classified by technology or the present way? Well, I think you have my answer on that. There is no way you can do it by technology, if I understand what you mean by that, and have the advantages that I'm talking about. They're not advantages only. They are essentials. They are absolute essentials. How can I determine whether or not I'm infringing if I have to just kind of wander through the halls hoping for the best? That doesn't make any sense. And D, ask about whether we should use different systems to improve our search. Well, that's a cost benefit question. Obviously, you want to improve if you can, depending on what the cost is. If the benefit and the costs are properly balanced, why, of course, you should improve it. But that doesn't mean you have any reason to abandon the system that we know about now. Concerning Question E, does the public need access to the examiner's search files? I'm surprised anybody would ask a question like that. If you mean specifically the examiner's search file, if you're in litigation, certainly you would. I mean today you can go to the examiner and normally he'll cooperate and tell you what his search file includes. But does this suggest that maybe people don't have to know what the examiner searched? I mean that's an absurd question. How are you going to litigate a patent if you don't know what the examiner searched, what classes he looked at? I don't know how you'd do that. It doesn't make any sense. For post-prosecution litigation purposes, does the public need to know what is in the examiner's search file? Again, how do you know whether this patent is properly examined if you don't know what he looked at? How could you possibly tell me that there's a presumption of validity when for all I know he or she may be looking out the window and pulling her side curls or something? That's silly. Of course you can't do that with an efficient system. In view of the current state of the art, including fully paperless patent and non-patent data bases, the question is once again how do I find out what the examiner did, how do I find out what is covered by this patent? In the end result, we have to have reliable, dependable system. I think the only way you can do that is in a claim based system that we have today. Thank you. MR. KAZENSKE: Questions? Mr. Banner, I have one question that knowing you many years and many of your colleagues that have continuously come to our attention, at least the past ten to 15 years, has been the fact that always the best art seems to be cited out of the European office. How do you account for that and why would that be if this is the best system? MR. BANNER: I don't account for that, and I don't think that's accurate. MR. KAZENSKE: Okay. What is your feeling of that, of the majority of the colleagues in the -- MR. BANNER: I don't know that a majority of the colleagues would agree to that at all. I think that most of the art that is--I don't even know what the basis for that assumption is. Seldom in my personal experience has better art shown up in European search results or European areas. Most of the time when you have patents invalidated--I think everybody is aware of the fact that it is, they are invalidated on the basis of prior art, which the examiner didn't know about or didn't search, on sale, public use, some patent that comes from someplace that the examiner may have not looked at a certain search class. That's the way they invalidat it. Foreign patents are important, of course, but they are not by any means determinative in the validity issues, of course not. MR. KAZENSKE: And my other question. You made a strong point under 282, the validity, and the point of foreign art being in the U.S. system. MR. BANNER: Yes. MR. KAZENSKE: On that, do you believe the current system as it has been and is maintained adequately puts that foreign art and has that foreign art maintained? MR. BANNER: You said now today if -- MR. KAZENSKE: And in the past. MR. BANNER: If it is true that the PTO no longer classifies new foreign patents in the search system. MR. KAZENSKE: Well, even over the past several, say ten years. MR. BANNER: Well, let me give you a full answer, if I might. If that's true, then that's in my view a real fault at the Office. A short time ago, the foreign patents were, in fact, classified. A long time ago, they weren't classified. They were hit or miss, and what should we do, I think, is the question? What should we do for the future? We should put the foreign patents that we know about, that would normally would go at a certain class or subclass, we should put them there so that if I as a member of the public or any member of the public wants to know what the prior art is, it's all front of me. Now, my problem is if those references known to the Patent Office are not put there, and I searched a certain place and I don't find the piece of art which the foreign art does, in fact, represent, I'm being deluded. How can I find out about that if you know about it, and I don't? MR. KAZENSKE: And if there is no record of what was there? MR. BANNER: If it isn't there, it isn't there. MR. KAZENSKE: Thank you. MR. TERAPANE: Mr. Banner, just to be clear on this, the USPTO, as you may know, is building a global data base which would utilize whatever indices are on foreign documents and including using foreign patent classification systems. They would not be classified in the USPC under one scenario of that global data base. Would that be adequate, in your view, to be able to search U.S. patents in the USPC and foreign patents in another classification system for many or most of the purposes that you need? MR. BANNER: I don't know what you mean by many or most of the purposes. I have to have a system which I can find out if I'm infringing a patent. Would that system do that? I mean this is not a question of what the text is. It's a question of what the claims are. MR. TERAPANE: Thank you. MR. KAZENSKE: Thank you, Mr. Banner. Our next speaker is Mr. Kenneth Gural. MR. GURAL: Kenneth Gural, Patentec, 2001 Jeff Davis Highway, Arlington, Virginia 22202. As owner of one of the leading U.S. search firms specializing in thorough investigations of the latest and most advanced technologies, such as complex electronics, I would like to state that I am strongly in favor of--let me repeat--in favor of the changes that have been proposed or already enacted by the USPTO, such as the desire to replace the U.S. Classification Process with mechanized classification and the abandonment of classification of foreign patents and other documents in the examiner's files and strongly in favor of any other similar changes that will make searching U.S. patents and other documents more difficult, cumbersome, time consuming, and therefore likely not to be performed by U.S. patent examiners because such changes will serve to make our jobs as private searchers more important, valuable and expensive. Already, it is a difficult task to locate relevant patents within the most rapidly growing subclasses generally within the most important and advanced technologies. These subclasses sometimes grow so fast that by the time a reclassification project is completed, due to limited funding, the new headings are already obsolete and contain many hundreds of patents. As a result, examiners already squeezed under the burden of quotas are forced ever more frequently to issue patents on non-patentable inventions for lack of locating the necessary prior art. As a private search company, that means validity searches. Knowing that patents on non- patentable inventions are issuing all the more frequently, companies are more prone to check the validity of issued patents. The companies hire us to complete the work started by the examiner. In other words, a sloppy classification system means lots of extra validity searches for private search companies. That is one reason why I am strongly in favor of the degradation of searching facilities at the USPTO. Additionally, a poor and sloppy classification system means that it will take longer to perform each search. That also means extra employment for searchers. The situation is rather like a well organized warehouse versus a sloppy junkyard. Knowing that a wrench lies in aisle C-5, bin 33, makes it obviously easier to find it than knowing that it lies somewhere out there in the vast wastelands of the junkyard. A team of searchers working 20 years might be needed to find it in the junkyard. A single searcher working five minutes may be able to find it in the organized warehouse. Ah, but let the computer find it. True. The computer can find a wrench, but can the computer find a bolt-tightening implement when told to find a wrench? It should be appreciated that a private search company would naturally seek any tool that would make searching more efficient. Otherwise, someone else would us that tool, and the search company would be out of work. We do use computers significantly everyday. But computers are only one of the many tools used during a search. And when we use computers, we generally use the combination of key words and classification information and less frequently use combinations of key words alone. In other words, the present U.S. classification information is extremely important, even in computer searches. Four, a computer used as a tool in searching is no different than any other human tool. The results will be in relation to the skill of the artisan wielding the tool. A plastering knife will produce smooth walls in the hands of an artisan and jagged lines in the hands of the inexperienced. Several vendors have proposed computer search products based on a concept search engine or method. These products have been hyped to the extreme, one vendor even claiming a few years ago that their computer program would make other searching techniques obsolete. That particular product made quite a stir when it was first introduced, especially because of such extravagant claims. Many companies tried it. None of those companies still use it. We tried it. The results were worse than a run-of-the-mill computer search using standard Boolean operators. Rather than make other search techniques obsolete, it seems the other search techniques made the search program extinct. In order to understand why, it's necessary to consider computer searching in general. Basically there are three steps to a computer search. The first is selecting the proper key words and proximity operators. The second is activating the computer to compute which patents fit that combination. The third is sifting through the output to filter out the irrelevant hits that inevitably occur in computer searching. For the concept-based search engines, step one, the selection of key words and proximity operators, is performed by the computer itself. At least, that's how it's supposed to be. In actual practice, for that particular product that I mentioned previously, the program required the operator to include so-called steering words to steer the computer to better results. I observed the program in action. At first the operator fed into the computer a short abstract of the invention. The computer being very fast almost immediately gave its results, a list of 50 patents ranked by relevancy. Among the highest scoring patents were patents having no relation to the invention. The operator reviewed the titles that were produced and selected two or three patents from the bottom of the list that seemed to relate to the general field of the invention. Then the operator provided the computer with a list of steering words selected from the two or three patents done in the first pass. The computer immediately gave a new list of 50 patents ranked by relevancy. Again, many of the patents at the top of the list were not relevant. Again, the operator provided the computer with additional steering words selected from newly found patents at arbitrary places in the list. After four or five iterations of this process, the list of steering words being over 20 words at that point was longer than the original abstract submitted for searching. Of course, these steering words looked suspiciously like key words to me except that with key words using Boolean operators and proximity terms, I generally know what I'm getting. With concept searching and steering words, I don't know what the algorithm is used that's deciding relevancy of the document. It's my belief that in computer searching, the first step, the selection of key words and proximity operators, and the third step, a sifting through the hits to select the relevant documents, are the most important steps, while the second step, the computation of which patents fit the key words, is less significant. Without going into further details about theory of computer search engines, suffice it to say that despite broad advertising hypes, there does not yet seem to be a computerized search engine so advanced that it can produce better results than an experienced computer searcher armed with a program offering standard Boolean logic and proximity operators. This is not to say there will never be a day in which a totally reliable automatic search engine will exist. That day will have arrived when a computer is able to properly understand and distinguish ambiguous phrases such as the "current control means," and this might appear in, for example, these sentences: switch 23 activates the current control means, thereby closing so and so. Or the current control means differs from the previous control means, blah-blah-blah. Or sentence three: eliminating the current control means there's no need for an adaptor. [Laughter.] MR. GURAL: Computers are not yet smart enough to distinguish these meanings nor will they be smart enough within the next ten years. But who cares about these subtle linguistic points? Isn't the purpose of searching just to enable an examiner to locate a few prior art examples in order to make a rejection? Who cares if another similar patent is overlooked because it uses somewhat oblique language that a computer did not properly understand so long as a few approximate examples can be cheaply located by blind computer techniques. Has anyone considered infringement, and actually several people have, but this was written previously so--the PTA seems to give the impression that its sole obligation to the public is to examine and issue patents. That's only half true. Or less. The patents that the PTO issues afford an inventor a limited monopoly in exchange for the disclosure of the invention. The inventor can use the patent to sue others who use the invention without permission. If I manufacture a new product, how can I be sure my product won't initiate a lawsuit? By doing an infringement search. By looking through the public records for claims relating to my product. The PTO never performs infringement searches. Therefore, the PTO may not even be aware of the differences between routine novelty searches and infringement searches. When we perform an infringement search, we are telling the company they are free to manufacture the item without fear of lawsuit or conversely that certain patents may be infringed if they proceed to manufacture. A single missed patent may result in millions of dollars of lawsuits. And the relevant portions of the patent that need to be searched are the claims, exactly the portions of the patent most likely to use broad and ambiguous language. In other words, exactly the portion of the patent most likely to be misunderstood by a blind computer. Presently, patents are classified in subclasses based on their claims. That is based on claims as understood by human, by an intelligent human. When an infringement search is performed today, it is generally limited by definitions of the subclasses, and at this point since my time is up, I'll say thank you. MR. KAZENSKE: Okay. Questions? MR. TERAPANE: Yes, I have one question. You know the Patent Office is fairly struggling with the degree or how to maintain the USPC in light of new tools that are available today that weren't available even five years ago, and I want to pose this question to you. Do you think that the USPTO is correct if they proceed down a path that they would modify how they reclassify or maintain this file in light of some marriage of new tools with the USPC? Do you see any need to change the level of support of the USPC? MR. GURAL: Well, to use computers to do the classification seems somewhat redundant if the same computers would be used to do a search later on. I mean why use, what's the point of having a subclass in the first place? The point of having a subclass is to have a human being look at the document and make a judgment relating to broad phrases that a computer can't properly parse and understand. Now was your question relating to classification or -- MR. TERAPANE: Well, it was more from the end use. Let's suppose that we arbitrarily say that whenever a subclass is a certain size, we'll reclassify that to break it down into a number of more manageable size subclasses. Let's say it's 150 -- MR. GURAL: That would be very commendable. MR. TERAPANE: Let's say it's 150 patents. In lieu of modern search techniques, though, could we say we'll leave a subclass of a thousand patents and say that would be one index for the patents that would be used in conjunction with some computer aided search that would enable the user still to do an effective and efficient search? MR. GURAL: Well, that already exists. There are many subclasses of over a thousand patents, and what we do for those is we either go through them by hand or we use a computer in conjunction with those large subclasses. But currently we have to pay for the computer time so is the Patent Office offering to make the computers available free of charge to the public to be able to do this? MR. TERAPANE: I didn't go that far. MR. GURAL: Oh, okay. MR. KAZENSKE: I'd like to have one question in view of your comments and the significance of the classification search and what you feel the resources that need to be put into that. Would you support the escalating cost of maintaining this system fee increases commensurate with the maintenance of that system? MR. GURAL: Fees to whom? MR. KAZENSKE: To users, applicants, a fee increase commensurate with maintaining that system and the cost of maintaining that system? MR. GURAL: If the costs are necessary, then someone has to pay for them. MR. KAZENSKE: And you would, I take that as a yes, that you would support fee increases? MR. GURAL: The question is to whom? MR. KAZENSKE: To the applicant or to a user of the system. MR. GURAL: Well, the bottom line here is we need effective search techniques. So whatever is most effective, that's what's needed despite the fact that the less effective it is, the better off we are. MR. SMITH: I have one. I hate to keep you on the hot seat there, but you seem to be very knowledgeable about the USPC and automated searching. To follow up on Mr. Terapane's asking about the size of the subclasses, do you think just in general, without getting into specifics, that there should be certain things that we could do, for instance, instead of breaking down subclasses, adding key words, to enhance searching in an automated environment and can get off the current traditional methods of reclass? MR. GURAL: Well, the main advantage of the reclass is the value added key words, if you want to call it that, that are applied to a document over and above the broad and ambiguous language that the applicant or his attorney would like to use such as the phrase "current control means" instead of "transistor." It doesn't really matter what form that value added element is. If it's key word phrases selected from a specific list of defined key words, you know, classification numbers, I mean to those who use the classification numbers those are almost equivalent, the numbers and what the subclasses mean, but the main feature is having a human look at that document and say this relates to the following subject matter, and organize it that way. Key words just are not efficient and effective for that except in a few cases, of course. MR. KAZENSKE: Thank you very much. Our next presenter is Mr. Eric Vacchio. MR. VACCHIO: My name is Eric Vacchio. I have an office in 2001 Jeff Davis Highway, Suite 505 in Arlington here. I'm glad to say I think most of my comments are probably going to echo what's already been said. And to an organization I think that even has to ask some of these questions in the first place, I'm glad to be repeating these same points because they're very important ones. My reading of the notice of hearing for this hearing had a very distinct tone to me of--your introduction talked about computerization of the Patent Office. To me, the notice of hearing had a very different tone than that, which was should we keep the classification as it exists or not? That to me is what I intend to address. And my answer to that is yes, we do need to keep the manual of classification, not only as it exists but we need to modify it as technology progresses. Several reasons for that: first of all, you have two groups of people that use that manual of classification. You have the examiners and you have members of the public, which each have different needs. As an examiner, for instance, I was an examiner in Group 1800. It was 180 at the time-- biotechnology. As an example, if an examiner is going to do a key word search in biotechnology, let's say he's working with the enviable molecule Interleukin 1. I bet there's not even five examiners who know that Interleukin 1 is actually called Interleukin 1 (IL-1), endogenous pyrogen (EP), leukocytic endogenous mediator (LEM), lymphocyte activating factor (LAF), et cetera. I'm sure that that's not even a complete list of how to refer to that biological molecule. A chemical example. If you have an organic chemical on your hands, you have the IUPAC method of naming that chemical. You have the Chem Abstracts method. You have the common name. In both of these examples, if you tell an examiner to rely on a key word, okay, he's got to sit down--first of all, he's got to know all the IL-1 terms. I've got the book that will teach it to him. If you need them, tell them to see me. And as I said, I'm sure that's not exhaustive. But when you tell the examiner to sit down and type these into the computer, what is he going to do? IL-1--IL-1? IL--in other words the word "IL" spaced one space away from the word "one". Is he going to have IL-1 together? Is it he going to have Interleukin1 with no space between the 1? If he truncates, if he decides to truncate, well, there happens to be IL-10, 11, 12, 14. He cannot truncate it. There is IL-1 alpha, IL-1 beta, sometimes with spaces, sometimes with not. In organic chemical terms, you might have a hyphen in the IUPAC name. CAS is not going to give you a hyphen. Maybe it will, maybe it won't. You have no way of knowing. It's incredibly difficult to do that when you have subclasses that exist, the manual of classification where a human has evaluated IL-1 or a human has evaluated, let's say, lymphocyte activating factor. If they're intelligent, more than the computer, they know to put it in the IL-1 subclass. That's what we need. We need a human judgment like that. That's a very specific, very, very realistic example of why you need the manual of classification. I'm not just saying this generally. There is tons of very specific examples like this. If you allow examiners to continue classifying like they do now in reclassification projects, an examiner then can always use the subclass as one of his key words and simply type in the subclass that he knows covers Interleukin 1. He will always get any patent in that subclass that calls it LAF or LEM. He's going to get it. And the only reason he's going to get it is because you allowed the reclassification projects to continue and the manual of classification to be as significant as it is today. I'm not even going to mention mechanical searches because almost no mechanical searches can be adequately searched on a computer. Those that can are a minority of the cases. The public is the other group of people that rely on this manual. the most significant point to me being, well, when the government gives an inventor a patent, what is he giving them? He's giving them exclusive rights, the right to exclude someone else from practicing his technology. That's a weapon. That patent is a weapon that can be wielded against anyone who might want to practice that technology. Fortunately, until now the government has somewhat balanced the power here. He gives the patent holder a weapon, yes, but in the form of the manual of patent classification, he has given other people a shield against that weapon. They know where to go and look to defend themselves against that sword, against that broad patent, and the only reason they know it is because the manual of patent classification exists as it does today. Key word searching--you cannot in order for an applicant for a corporation to go to market with a product, they have to do an infringement search. They have to use, and that infringement search absolutely hinges, it's critical, it's not dispensable at all to the manual of classification. That infringement search has to use it. If we try--the only alternative is to try a key word search. Well, patent claims, as everyone in here knows, are drafted to be broad. A fine example of the wrench turning implement or nut turning implement is a wrench. I could give you countless biological examples which get even amazingly complex to describe a molecule that may have never been seen on the face of the earth before. There are no words to describe it other than it has a certain molecular weight, it has a certain glycolysolation pattern, it has a certain activity, it's inactivated in this circumstance. That's how many biological claims read. You cannot key word search any of that, none of it. Even if there is a concrete term that's known and exists in nature within the patent community, claims again are drafted to be broad. They are not going to include a limiting word in their claim, an attorney drafting a claim, unless he has to. If there is any way at all to come up with a broader claim, he's going to do it. So I think that just about let's you know what I think of it. It's not a question. This hearing should not even be being held with that as one of the posed questions because you're removing the foundation from the building here and asking us to come up to the ninth floor. No thanks. We can't do that. And I think that's about all my comments are going to be. MR. KAZENSKE: Questions? MR. SMITH: Do you have any experience with other classification schemes or systems besides the U.S.? MR. VACCHIO: Besides the U.S. manual of classification, only the IPC and not to have searched it but only as an examiner you need to come up with IPC equivalent subclasses when you classify a U.S. patent. I've never done a search under that system, no. What I do know it's significantly smaller than our system right now. And if we were to shift over to let's say IPC, you're going to double or triple the number of patents in any given technology group or subclass than we have. Now, your examiners, they can't do a proper search as it is with the time they have. You want to double or triple the size by going to a different system, they're not even going to look at the subclasses anymore, let alone search them. MR. TERAPANE: Eric, you did point out, though, that it was useful to be able to use the classification designation as a key word and marry that with other key words. So that seems to suggest that again going back to a previous comment that I made with the previous speaker, that you could have a large subclass that could be effective if married with other key word searches? MR. VACCHIO: No, not for infringement searching and potentially not for validity searching. For a patent examiner maybe. If he's doing a patentability search and he wants to rely on a very specific key word, it could be useful. But as I said, there are probably less than five examiners who know that lymphocyte activating factor is actually IL-1. MR. TERAPANE: Well, they wouldn't have to know that if they were using class/subclass right because all the patents would already be in there? MR. VACCHIO: That's true. That's true. But to use that patent subclass requires that a reclassifer used his brain on it first. For infringement searching, no, it's absolute. You cannot do an infringement search even by combining key words with a class/subclass term because that, you have to zealously represent your client, and that is not being zealous at all. That almost borders on malpractice because you cannot limit and sit there and guess that you can judge what broad terminology might exist in that large subclass. I know I wouldn't jeopardize my clients that way. MR. TERAPANE: Thank you. Just one other question. Did you identify who you were representing? I may have missed that. MR. VACCHIO: I represent myself although I am a private searcher. So I have the majority of the searching that I do is infringement and validity work. And it's from eight years of that experience that I make these comments. MR. TERAPANE: Thank you very much. MR. KAZENSKE: Thank you. The next presenter is Mr. Jonathan Grant. Mr. Grant is not here. The next then on the list is Dr. Samuel Shipkovitz. Not here. Next on the list is Mr. William Plunkett. MR. PLUNKETT: I'm Wilson Plunkett, 11609 35th Avenue, Beltsville, Maryland, and I'm here representing the Patent and Trademark Depository Library Association, and I'll be reading two very short letters to you today. The first is from the president of the PTDLA, Mr. Robert Williams, of the Anchorage Patent and Trademark Depository Library. I am writing to you to express the concerns of the Patent and Trademark Depository Library Association regarding proposed changes in the United States Patent Classification System. Ours is an organization of librarians who are charged with providing patent search services to the general public. Most of us work primarily with independent inventors with little or no experience in patent searching. All of us rely upon the USPS as we attempt to provide accurate and complete search services to the public. The USPCS is perhaps the most vital tool used by all patent searchers. It is certainly the most important tool used by lay searchers of patent materials. The USPCS is the most organized and most precise retroactive patent search tool in existence. The precision comes in part for more than 440,000 subclasses compared to the mere 149,000 subclasses in the IPC. The information in the USPCS is cumulative while in the IPC it is not. So that if you're doing a search in the IPC, you have to perform multiple searches in order to trace classes, subclasses and reclassifications all the way back. And even then, you can only go back for 30 years. Even professional patent searchers would be hard-pressed to do a thorough patent search using only those tools. While it is true that a patent classifier would be able to process many more patent applications that way, it appears that the workload of the examiners would be greatly increased. Where is the economy there? Reliance on a group consisting of commercial data base suppliers and domestic and private entities to provide a hodgepodge of classifications, abstracting, and summarization work appears to be sheer folly. We have seen many governments and private provider files disappear from the Internet and from commercial data bases. This appears to be another example of a case where a governmental entity is necessary to provide standards and guidance for such complex and vitally important a system. The U.S. Patent and Trademark Office may wish to consult with the Library of Congress regarding the development of classification systems to handle large bodies of knowledge. The Library of Congress developed the ALC classification system in response for the need for specificity. Melvil Dewey had earlier in the 1900s developed a classification system that's still used by many small public libraries and school libraries that divided all non-fiction books into just ten broad classes, and used additional breakdowns within those classes. The Library of Congress tried numerous systems over the years trying to provide access to this huge volume of information. But the LC's classification system as it stands today allows you to divide individual classifications into subclasses and you can expand the system as it becomes necessary. Libraries all over the world use the LC system because it's the best system for the job. Because it groups like terms together, a researcher frequently can discover related but unexpected avenues of research. The Library of Congress and the U.S. Patent and Trademark Office provide the kind of services and leadership that can only be provided by a large governmental entity. Many people who need patent information do not have access to the internets or on-line data bases. And they do not possess the skills necessary to perform sophisticated searching. The old saw, "If it ain't broke, don't fix it," has never been a more appropriate comment. I urge you to give very careful consideration to any changes that are proposed in the United States Patent Classification System. Thank you for the opportunity to provide testimony at this very important public hearing. Sincerely, Robert C. Williams. The second, much shorter letter, is from one of the patent depository librarians out in the field. It's from Dave Morrison. He's at the Salt Lake City Patent Depository Library. I'm writing to express my concern regarding the notice of hearing published by the PTO in the May 2, 1996 issue of the Federal Register. The Marriott Library and the University of Utah is the Patent and Trademark Depository Library for the state of Utah and for nearby parts of the intermountain west. I am the librarian responsible for the PTDL collections and activities. I believe that it would be a terrible mistake for the PTO to stop the maintenance of the Patent Classification System. My patrons, mostly independent inventors with little or no experience searching computerized data bases, rely upon the classification system to make certain they've examined all the areas of technology covering their inventions. Without the authority of the classification system to establish the scope of their searches, my patrons will be far less certain that they can proceed with their applications. The articles that I have read concerning the abandonments of the USPCS have convinced me that no other patent classification system would be a satisfactory substitute. The International Patent Classification System, which is specifically mentioned in the Federal Register notice as a possible replacement for the USPCS, has serious deficiencies as a search tool. First, the system is only revised every five years. This guarantees that it will never be a state-of-the-art system. Second, no effort is made to reclassify older material into the new classification system. So there are, in effect, six separate data bases that must be examined. And third, IPC categories have titles but no definitions of contents as are seen in the USPCS. As a result, the document placement in the IPC is seen as less consistent than in the USPCS, which means a given technology may be split into different classifications. Thank you for this opportunity to comment on the PTO's notice. David Morrison, Marriott Library, University of Utah. Then I had a very short comment of my own. The concern here today seems to speak to the idea that full text searching or key word searching cannot work. Well, I want to let you know that I've seen it work. I was at Baton Rouge Patent Depository for four years. Every year we had second through six graders come in to do patent searches and key word was completely sufficient for their needs. [Laughter.] MR. PLUNKETT: Perhaps there weren't any great economic impacts, though, if a seven or eight year old failed to find some prior art. Thank you. MR. KAZENSKE: Questions? I have a question. First of all, I'd like to make a comment. According to the mandates of the United States Congress, this agency does not receive public money such as the Library of Congress does, and we operate on our fee revenue. But following a comment that was made earlier, if you were not here, there was a comment that maybe the PTDL program would support a revenue to maintain this vital system, as you so do. Do you favor fees in one form, and it could be in many factors or a revenue, to maintain that system from the PTDLs? MR. PLUNKETT: Speaking only for myself and not for the association, because I don't have their input on this, I believe that any fees that are necessary to maintain the system could certainly be applied. At the same time, the Patent and Trademark Depository Libraries provide a service for the Patent and Trademark Office by in a certain sense giving them 84 branch offices spread throughout the country and making it very possible for small inventors, small business entities, to perform at least an initial part of their patent search in the area where they live, particularly if they cannot afford to come and use the public search rooms in Arlington. MR. KAZENSKE: Thank you. That is the entirety of those that had pre-signed up. We have two individuals that signed today in the back of the room, and I have that list. I'd like to give them that opportunity now. The first one is a Rodger Flagg. MR. R. FLAGG: Thank you, sir. My name is Rodger Flagg, and I'm president of Express Search. We do patent research for attorneys and agents all over the country in excess of 1500 a year. I personally review each search and sign off on them. I'm also a registered patent agent. I passed the patent bar in '78 and have been practicing since. I also am an inventor with 12 patents to my name. First of all, I think the majority of the speakers have fairly addressed the questions that you've asked, and I would just like to add a couple of points. First of all, an inventor by definition can be his own lexicographer which means he can choose his words to define his invention. And those words are not uniform, and that is the result of the problem with most computer searches. What is a lever to me may be an actuator to you and may be a handle to you. If you don't put all the terms in, the patents that use that alternative terms are going to be missed in the search. So a computer search while it's helpful and we use it in every search as a supplement to our manual search, it is not nearly as comprehensive as the combination of a computer search and a manual search, and routinely--another point I want to make--routinely in the majority of patent searches that we do, we find relevant art more than 30 years old. You perhaps can give me the name of the Patent Office commissioner back in the 1880s that said everything worth inventing has already been invented, let's close the Patent Office and save the taxpayers some money. He certainly was wrong in that regard, and Congress was wise in not doing that. At the same time, there's a lot of prior art that exists more than 30 years back that is very relevant and is necessary to review, and it can best be reviewed in the current system. The inventor is charged with having prior knowledge of all relevant prior art. And that is a challenge that makes the searching essential to filing an application. He needs access to that art, and our system while it's certainly not perfect is the best that I know of out there anywhere, and this is repeated to me by people that have used the other systems. I searched the CSIR, the APS, the CASIS systems. I don't get into the foreign. But it's been mentioned already that there are like 140,000 subs in the system you're talking about and about --140,000 in the other system--440,000 or so in our system. This narrowing of this classification is significant for doing an efficient search. If we group it in much larger categories, what happens as a result is it takes a lot more time to do the search. This may be more efficient to the examiners, but it's certainly going to add costs to the inventors, significant costs. And the likelihood, as a result, is that they will not do novelty type searches which means that the examiner has to work harder. A well-done preexamination search is a great aid to an examiner in forwarding the prosecution of that application. And by having a preexamination search, the attorney or agent can prepare the application to focus on the novel features of the invention, something that is not possible if you don't have that knowledge of the prior art. If you're writing a patent in general terms, it's much harder to prosecute. So where you're saying on one hand, you may be losing on the other. Most inventions fall in the cracks between two other inventions, and this is where the classification system comes into play, until there is significant art there to develop their own subclass, these are in two separate locations. This--let me give you an example--when lasers-- technology always leads terminology, and terminology is what computers search. Technology leads that. When lasers were invented, the term laser didn't exist. That's an acronym that was developed after the fact, and there are a number of significant laser patents that do not use the word "laser." Now if you put the term laser in the computer, you're eliminating that important prior art, and they may not have even been brought forward into the current classification of lasers. So you need to be able to go back and reconstruct what existed at the time of the invention, at the time of the technology prior to the time of the terminology, and our system does that better than any other system we can have. So I'm strongly in favor of continuing the existing technology and with our classification system. I think the Patent Office is doing a great job and I'd like to see them continue in that effort. Thank you. MR. KAZENSKE: Thank you. One question. MR. R. FLAGG: Yes. MR. KAZENSKE: You made a comment that should this not be maintained, it would be a cost to the inventors. Do you have any idea of what that may be in real dollars? MR. R. FLAGG: Yes, they were talking that an examiner in one of your earlier testimonies in using the foreign system was like 17 hours to do a search and we do it in four to five hours. That difference is a multiplication of three or four times. If the search cost costs three or four times to the inventor the amount that he currently pays, a lot of people are going to opt not to do the search with the result that it is going to be a handicap both to him as an inventor and to the examiner in prosecuting that case, not having the availability of that professionally prepared search as a starting point for his further review. MR. KAZENSKE: Having asked this question to many before, and receiving the answers I've received and maybe the non-answers I'll give you the same opportunity to evade the answer also. At the increased cost of maintaining this system, would you support fee increases commensurate with the continual cost to maintain the system? MR. R. FLAGG: I heard your question before, and my question is where do these fee increases come? Do they come in the cost of filing an application? MR. KAZENSKE: They could. MR. R. FLAGG: Do they come in the cost of copy? Right now we pay 25 cents a copy for everything we copy. Do they come, you know, in what form do these fees increases come? And the other point--I'm glad you asked that--because I had one more point I wanted to make--in basic economic classes, you understand that you can talk about anything in terms of a pie. And if you have a pie, and you want to increase your revenue, you have some options. You can either increase the pie or you can take a bigger part of it. If the Patent Office is moving in the direction of increased fees for the examiners and reducing their level of efforts, these are not producing more pie. These are increasing the costs without producing more results. If that's true, then we're asking the inventor to share that greater burden for his right to restrict others for a limited period of time. MR. KAZENSKE: Thank you. And your last thing I'd just like to comment on--your quote--I've done some research on this myself. It's attributed to the Commissioner Duval in 1898. MR. R. FLAGG: Okay. Thank you. MR. TERAPANE: Mr. Flagg, I enjoyed your comments, and I don't quarrel with the numbers, the relative numbers you were talking about, but for the record, I just want to clarify the numbers. There's 140,000 subclasses in our system and 60,000 in the IPC. So the ratio was about right. MR. R. FLAGG: Okay. Excellent. Excellent. MR. TERAPANE: I just wanted to clarify the record. MR. R. FLAGG: Another way to look at that--excuse me, sir--in just finishing your question there, if you look at the face of any patent, you can see the number of international classifications versus the U.S. classifications. You can see routinely that the U.S. classifications are more defined, more distinct, and this is a great aid in searching in validity, infringement as well as novelty reexamination. Thank you. MR. KAZENSKE: Thank you very much. The last individual that has asked to comment today--is it Cristopher Flagg? MR. R. FLAGG: That's my son. MR. C. FLAGG: Hello. My name is Cristopher Flagg. I'm a private patent searcher, and I've worked in the public search room for ten years, five of those years doing mechanical patent searches. My address is 123A Longwood, Charlottesville, Virginia 22903. And as I said, I represent myself in saying this. I'd like to start out by saying that I have dealt with the IPC some. I've not actually searched it, but I have gone through it on several occasions. And what has not been mentioned so far in relation to the examiners is the level of expertise that will be required under a less broad system. A lot has been said about the increased amount of time that examiner's searches will take, but in a broader system with more general classifications, the examiners will be required to understand more. Currently, under the system, each subclass, there are a set of examiners who know specifically what is maintained in that subclass. And for any particular set of subclasses, there are several examiners. To expand the size of the subclasses to reduce by say a third would cause the examiners to be responsible for three times the intellectual knowledge contained in those subclasses, and this is not berating the examiners--they have a hard enough time as it is just knowing what's in their subclass--a subclass three times as large will cause them to know more but not be experts in any of this. As a result, as the examiners do their searches on the applications, the level of integrity of those will go down. Unfortunately, as a result of this, this will cause the actual validity of patents to be determined more at the post-prosecution litigation stage which is more costly. When you're looking at the cost to the particular inventor as has been mentioned often, an increase in fees initially would be difficult, but to broaden the system, the result would move from an application based system to a litigation based patent system which is infinitely more expensive. Second, I would like to address the issue of the computer searches. It's been stated several times that key word searches are not sufficient, and I would like to second that. They are essential, however. They are used in my search strategy, and I use them on every search I do. And they are a supplement. They are a much needed supplement, but they are only a supplement. To remove them would add a burden to the searchers, but the searchers cannot rely on that supplement alone. I know several systems have been proposed, and I think adequate explanation as to why those don't apply has been given. I can see two possible scenarios for what is happening. First of all, the Patent Office must either go for complete harmony. If we're talking about harmonizing with the European system, their system is also changing as is ours. I would assume that their system is not static with the number of classes, et cetera, that they have. So for us to try to change to them will be a continual uphill battle. If there is some plan for global harmonization of intellectual property, such as a consortium or an intellectual property standard set, then I have no qualms about moving to a different system under that in the means of harmony. But until that point, if we are still dealing with a system, our system is, as has been stated, one of the best. Second of all, as to the approach for computer searches, the first and foremost problem with the way it's being approached is that several companies are taking their ideas and making tools. Tools are wonderful, but tools should be generated as they are needed. What I foresee as a possibility to the solution of this dilemma is to create an open-ended Patent Office data base where the patents are presented and it is up to the general public to come up with tools. Now, again this would be just a supplement to the way searches are done now. But by having an open-ended system, you would have a more market driven generation of tools as opposed to now where we have technology, let's generate a tool. The purpose of tools is to aid. If the question is every asked we have a tool, what can we do with it, that tool should be considered useless. I'd like to close in saying that the way in which this hearing was presented led me to believe that there were two questions being asked. The first question is if we move, when we move to a paperless system, should we move to a new type of classification? The second question was when we move to a new type of classification, should we move to a paperless system? And these two were asked in conjunction which leads me to believe that it is prior to this hearing the plan of the Patent Office to move in both of those directions. I hope that the discussion here today has given some insight into the feelings of not only the examiners but the lawyers and the public searchers along with the inventors dealing with this. Thank you. MR. KAZENSKE: Thank you. I have no question. You made quite a statement about how the examiners view this. Were you a previous patent examiner? MR. C. FLAGG: I was not a previous patent examiner. I'm sorry if my statement led it to seem that I was saying the examiners felt like this. It was my understanding of having dealt with examiners. MR. KAZENSKE: Thank you. MR. TERAPANE: Cris, just a comment. You know it's recognized up here that the concepts of paperless and indices or classification systems are different concepts. It is also recognized, though, and I think we all should appreciate this, that as long as we have a paper-based system, we're limited on the number of indices and what we can do to augment that file to aid retrieval. And so as we move into a paperless environment, and this is part of what the future holds for us, we're going to be able to put virtual classifications on documents that would be impossible in a paper-based environment where you have to create space in a file in a cabinet and support to maintain that. And that's part of where we are today. MR. C. FLAGG: Yes. MR. TERAPANE: So as we move into a paperless environment, you know, what are the indices that are going to be used for searching and what are the possibilities and potentials that don't exist today that will exist tomorrow? MR. C. FLAGG: If I may make a comment on that? MR. TERAPANE: Yes. MR. C. FLAGG: I agree with you wholeheartedly that a paper-based system is a sinking ship, and that unless we move to upgrade our technology that this ship will be sunk in not a short time but will sink. However, I would feel more comfortable as a public searcher to have a dock in place for this ship to go, and that would be a previously established system, a system that is known to work and can viably replace the paper system with a paperless system rather than what it seems to be now which is jumping off ship and swimming to shore hoping that we can find a driftwood system along the way. Thank you. MR. KAZENSKE: Thank you. MR. TERAPANE: For somebody who is not a great swimmer, I can appreciate that comment. Thank you very much. MR. KAZENSKE: Thank you. Let me thank everyone that came here today, took their time to give us those comments. Believe me they go with a lot of appreciation. I will make a comment. There's been a lot of discussions here about what an examiner feels and that. I do not come to this position without knowing that. I've been a patent examiner in this office for 14 years, plus many other positions in this office. So I don't come into this position as a total vacuum of what it is to examine a patent application. With that, let me just close. This completes the list of persons who have asked in advance and today to testify. Let me mention that the written comments on the Notice of Hearing and Request for Comments on a Technical Documentation Strategy must be submitted by Friday, June 7, 1996. A transcript of the hearing will be prepared and will be available along with any written comments received for public inspection in Room 902 of Crystal Park Two, 2121 Crystal Drive, Arlington, Virginia. Both the transcript and the comments should be available by June 14, 1996. In addition, the transcript of the hearing will be available on the Internet through anonymous file transfer protocol, address: ftp.uspto.gov. All written comments and oral testimony will be taken into consideration in the implementation of any administrative changes in the Patent and Trademark Office related to this issue. Any written comments received after June 7 cannot be assured of consideration. This concludes today's hearing. Thank you very much for your testimony and your attendance here today. Thank you. [Whereupon, at 11:45 p.m, the hearing was adjourned.]