Research on hot news classification algorithm based on deep learning

Abstract

Nowadays, the Internet development in full swing, has given rise to a variety of network applications, network news as the basic application in the information age, has been more and more customers favor, however, news website to report news type number many, every day the user focus on hot news, so I need to forecast the hot problem in the news, so in view of the study of this problem, this paper studies deep learning application in text categorization, combined with the characteristics of the news text, put forward the double Bi-Gated Recurrent Unit(GRU) + attention deep learning model to predict hotspots, and achieved good results.