| 摘要 |
PROBLEM TO BE SOLVED: To provide a method and a system for classifying a display page, based on an automatically generated summary of the display page. SOLUTION: An Web page classifying system generates the summary of an Web page, using an Web page summarizing system. The summary of the Web page may sometimes include a sentence of the Web page related most closely to main topics on the Web page. The summarizing system identifies the sentence of the Web page expressing the main topics on the Web page, using combiningly advantages in a plurality of summarizing techniques. When the summary is generated, a classification system applies a conventional classification technique to the summary to classify the Web page. The classification system uses the conventional classification technique such as a simple Bayesian classifier and a support vector machine, and identifies the classification of the Web page, based on the summary generated by the summarizing system. COPYRIGHT: (C)2006,JPO&NCIPI
|
