A Short Text Classification Algorithm Based on Semantic Extension

A semantic-extension-based algorithm for short texts is proposed, by involving the Word2vec and LDA model, to improve performance of classification, which frequently deteriorated semantic dependencies scarcity features. For every keyword within a text, weighted synonyms related words can be generated Word2Vec respectively, subsequently inserted extend text reasonable length. We not only have established criterion means similarity estimation determine whether sentence should extended, we designed scheme choose number extended words. The will classified. Experimental results show that, classification proposed algorithm, in terms precision rate, approximately 5% higher than that TF-IDF model 10% VSM method.