An Improved Video Recommendations Based on the Hyperlink-Graph Model
Wenqian Shang, Songtao Shang, Shuchao Feng, Minyong Shi · 2016
The graph-based algorithm for personalized recommendations mainly depends on the user-item model to construct a bipartite graph. We can provide recommendations by analyzing the bipartite graphs. However, for personalized videos recommendations, the classical recommendation algorithm based on graphs has low efficiency. Therefore, this paper gives an improved video recommendation algorithm that is based on the hyperlink-graph model. This method cannot only improve the accuracy of the recommendations, but also reduce the running time. Furthermore, the Internet users may have many different interests. For example, they may be interested in sports videos, and at the same time also enjoy watching political videos. For this reason, we propose a complement algorithm based on hyperlink-graph model for video recommendations. This algorithm is based on the principles of min-Hash. It realizes the cross clustering in user layers and further improves the accuracy of personalized video recommendations.