文本分类（Text classification）.doc

文本分类（Text classification） Definition of text classification What is text automatic classification: Use a computer to automatically categorize text sets (or other entities or objects) according to a certain classification system or standard Automatic classification based on classification system Automatic classification based on information filtering and user interest (Profiles) Automatic classification based on classification system The classification system is the classification of the word statistics Keyword classification, now full text search The correct cut of the word is not easy to distinguish. Automatic classification based on information filtering and user interest Learn the knowledge and strategy of humans classification of text The correlation between a persons correlation between text and category is learned from the association between a file and a tag category Edit the text categorization process Generally includes the expression of the text, text categorization classifier selection and training, the classification results of evaluation and feedback process, including the expression of text and can be subdivided into text pretreatment, indexes, and statistics, such as feature extraction step. The overall function module of the text classification system is: (1) pretreatment: the raw material is formatted into the same format for subsequent unified processing; (2) indexing: breaking down documents into basic processing units while reducing overhead for subsequent processing; (3) statistics: word frequency statistics, related probabilities of terms (words, concepts) and classification; (4) feature extraction: extract the characteristics that reflect the theme of the document from the document; (5) classifiers: the training of classifiers; (6) evaluation: analysis of test results of classifiers. Edit the text classification method Text categorization problem there is no essential difference with other classification problem, the method can be summed up i