Index term extraction apparatus surveyed document, personality expression diagrams, and document feature analyzer

摘要

The first appearance frequency calculating means for calculating the index words and extracting index terms extraction means, the extracted index words surveyed document d in the function value IDF frequency of occurrence in the comparison target document group P (P) is , and similar documents group elected means that based on the data of the survey document d, to elect a similar document set S that is similar to the survey document d from the comparison document group P, and index terms that have been the extraction, similar document group and a second appearance frequency calculating means for calculating function value IDF frequency of appearance in the S a (S), on the combination of the function value of each frequency of occurrence for each index term, in a similar set of documents and comparison group of documents is the calculated Based on, and an output means for outputting the data of the time series variation and its positioning and its respective index words. As a result, it is possible to have accurate understanding of the temporal transition and nature of the survey document without reading the document.