A News Recommendation Algorithm Based on Word2vec and Convolutional Neural Network

Abstract

The information overload of news makes it difficult for users to find news they are interested in. How to obtain the news that users are interested in among tens of thousands of news has become an urgent need in the current news recommendation field. Therefore, this paper proposes a news recommendation algorithm based on Word2vec and convolutional neural network. Firstly, the news content is modeled, and Word2vec is used to train the news word vector model, and then a convolutional neural network model is used to classify news; secondly, the user interest is modeled to obtain a user-news topic preference matrix; finally, a collaborative filtering algorithm is used to recommend news based on the user-news topic preference matrix. The experiments show that the news recommendation algorithm based on Word2vec and convolutional neural network has better recommendation performance.