摘要 |
PROBLEM TO BE SOLVED: To provide an automatic classification system and method for efficiently classifying a document to a proper category according to the content, and for easily realizing retrieval. SOLUTION: This system is provided with a feature learning means 20 for learning the feature data of classified categories, classifying means 31 for easily classifying a document to plural categories in the range of plural classification constraint data 8 according to the feature data of each obtained category, and classifying means 32 for precisely selecting only the optimal category and classifying the document. Also, this system is provided with a switch 4 for selecting any classifying means according to the category to which the document to be classified belongs.
|