Step:        Input:          Process:            Output:

       1    Document(s)             NLP               TermsD0~

                                                      "Parsed-Doc"


       2        Topic(s)            NLP               TermsT0~


       3

                            Hand Filter:
              TermsT0~      1. "Eliminate"            Weighted-TermsT0~
                            2. "Weight"-3/2/1

                                                    "Source- Query"

                 Figure 6: Schematic R~presentation of Data Preparation


WS3891102-0187


"ucdersott" na Rcdermott
"international" adj international
"inc." ukw? inc.
"sajd" vt-past say vt-pastprt say
"its" gen its
"babcock" na babcock
"\&" *and* and
"wilcox" na wilcox
"unit" Sn unit
"coRpleted" vt-past couplete vt-pastprt couplete
"the" d the
"sale" sn sale sn sell
"of" prep of
"its" gen its
"bailey" sn bailey
"controls" vt-pressg3 control pn control vt-pressg3 control
"operations" pn operation

"of" prep of
"about" prep about
"$370" ukw? $370
"uillion" quant uillion
"\." *period* \.


            Figure 7: Sample of Data-Document After Morphological Analysis


                                   259