An Improved Classification Method of Expanding Feature Vectors of Short Text Words

Abstract

The present invention relates to an improved method for classifying extended short text word feature vectors. The method introduces the Word2vec language model to carry out word embedding for short texts, expands short text feature vectors to solve the sparsity of short texts, and converts word vectors into probabilistic semantic distributions. To measure the semantic relevance; for the extended feature vector of the short text, use the improved feature weight algorithm and introduce the semantic relevance to process the expanded word feature vector. This method can distinguish the importance of words in the expanded short text to obtain more accurate semantic correlation, which can effectively improve the classification effect of the short text. Experiments prove that the method in this paper can ensure the accuracy of short text mining feature vectors, and at the same time greatly improve the accuracy of short text classification, and the effect is remarkable. It can be used in decision-making directions in various fields such as hot topic classification and mining, monitoring public opinion information, etc., and has strong practical value.

Description

technical field [0001] The invention relates to an improved classification method for expanding short text word feature vectors, in particular to an improved classification method for word embedding and extended short text word feature vectors based on a Word2vec model, and belongs to the technical field. Background technique [0002] With the rapid development of social networks and e-commerce, short text forms such as Weibo, Twitter, product reviews, and real-time news push have become the mainstream content of the Internet. Short text is generally defined as shorter in length, ranging from 10 to 140 words. Research on the classification and mining of hot topics in short texts and the monitoring of network public opinion information have important application prospects for decision-making in various fields. Therefore, how to efficiently and correctly mine short texts has become a hot research direction. [0003] For conventional text classification, most of them use the t...

AI Technical Summary

Problems solved by technology

[0007] The object of the present invention is: for the defect that prior art exists, propose a kind of classification improvement method of extended short text word feature vector, use the word2vec technology in the neural probabilistic language model to carry out word embedding to train the word vector in the extended short text, with Solve the technical problem of fewer short text feature vectors

Images