Improved SIFRANK for Efficient Media Hotspot Mining in Social Networks

Abstract

In the era of information explosion, social media has become the main platform for the public to obtain information and express their opinions. How to quickly and accurately mine media hotspots from massive data has become an urgent problem to be solved. With the rapid development of social media, media hotspot mining technology is facing higher requirements. This study focuses on improving the SIFRANK algorithm and proposes a more efficient and accurate method for mining social media hotspots. By deeply mining the emotional tendencies and interaction patterns of social media users, as well as introducing information timeliness evaluation and optimizing network weight calculation, the improved SIFRANK algorithm significantly improves its performance in hotspot recognition. Tested on the Twitter dataset, the improved algorithm achieved a 15% increase in accuracy in identifying hot topics, reaching a 92% accuracy rate (compared to the baseline method of 77%), and was able to respond more quickly to newly emerging hot events. In dealing with complex network structures and changes in information propagation speed, the algorithm has also shown stronger adaptability and robustness, with a 5% improvement compared to traditional models such as PageRank. This study, through technological innovation, not only improves the efficiency and accuracy of hotspot identification, but also provides a powerful tool for understanding social public opinion trends and guiding public policy formulation.