| 摘要 |
PROBLEM TO BE SOLVED: To identify and treat a document not sufficiently fit to any of class groups in a model obtained by automatic classification or clustering. SOLUTION: A probabilistic classifier or categorizer 20 generates the model for associating each document with a class, by performing probabilistic clustering or probabilistic categorizing of the plurality of documents. An outlier measure calculator 32 computes outlier measures for the documents indicative of how well each document fits into the model. An outlier thresholder 34 identifies an outlier document for a user, based on the computed outlier measures and a user selected outlier criterion. COPYRIGHT: (C)2009,JPO&INPIT
|
