Text classification incremental training learning method supporting vector machine by compromising key words

Abstract

The present invention belongs to an intelligent information processing technology, in particular, it relates to a text-sorted incremental training and leaming method by combining keyboards to make learning and support vector machine. The invention utilizes the important action of keywords in the training process to provide the invented synchronously-regarded keyword "incremental" learning method, in the incremental training process the learning and regulation of sorted keyboards can be made simultaneously, so that said incremental learning method can obtain the sorted accuracy identical to that of one-step training process.

Description

technical field [0001] The invention belongs to the intelligent information processing technology, and further relates to the text classification processing technology, in particular to a support vector machine text classification incremental training and learning method integrated with keyword learning. Background technique [0002] With the rapid development of network and information technology and the great abundance of digital document information, the classification and processing of texts, materials, web pages, etc. has become an important technical means of information processing. For text classification, the support vector machine method (support vector machine, SVM) is currently one of the most effective methods, in 1998 Joachims in the document "Text Categorization with Support VectorMachines: Learning with Many Relevant Features" (In Proceedings of the European Conference on Machine Learning, Berlin, Springer, 1998) verified its excellent performance in text clas...

AI Technical Summary

Problems solved by technology

[0004] This method is aimed at the phenomenon that in the current incremental training of SVM text classification, because only the optimization of the support vector itself is considered, the effect of incremental training is slightly lower than the classification accuracy of one-time training. Combining incremental training and one-time training in In order to eliminate the difference between the two, the incremental training is consistent with the one-time training. The purpose of the classification accuracy

Images