Research on Keyword Extraction of Word2vec Model in Chinese Corpus
Chenchen Zhang, Xingjuan Wang, Shuiyuan Yu, Yongbin Wang · 2018
With the rise of neural networks in recent years, Word2vec, a tool based on neural networks, has been widely used. Word2vec, as a new technology for natural language processing, whose appearance has made word expression in computers a major breakthrough. At the same time, more and more people are keen on sports news because of the development of national fitness exercise in our country. However, the existence of professional jargon in the news makes it obscure to ordinary readers. In this paper, Word2vec was applied to sports news, and was used to explore Chinese corpus and keyword extraction. The experimental results show that Word2vec can match related words in descending order of similarity so that readers can understand the news better.