Open AccessA Hybrid News Recommendation Algorithm Based On K-means Clustering and Collaborative Filtering
Author(s) -
Jing Liu,
Jinbao Song,
Chen Li,
Zhu Xiaoya,
Ruyi Deng
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1881/3/032050
Subject(s) - collaborative filtering , computer science , cluster analysis , recommender system , process (computing) , popularity , data mining , algorithm , the internet , information retrieval , artificial intelligence , world wide web , psychology , social psychology , operating system
In the era of paper media, the information channels and the information content were integrated. With the birth of the Internet, they tended to be separated while the information channels continued to expand, which brought a massive amount of news information to process. Therefore, it’s essential for us to adopt new methods and new models to deal with all the information. This paper gives a brief overview of news recommendation technology, and proposes a hybrid news recommendation algorithm, which combines content-based recommendation algorithm and collaborative filtering, using TF-IDF method and K-means clustering technology to extract and process the features of news content, meanwhile, this paper applies SVD technology to solving the matrix sparse problem in the traditional collaborative filtering algorithm. Moreover, news popularity is taken into consideration in this paper then it combines the candidate recommendation results of each approach. At last, this algorithm achieves a better result compared to traditional recommendation algorithm’s result.