Exploiting Information Diffusion Feature for Link Prediction in Sina Weibo

Area of Science:

• Social Network Analysis
• Information Diffusion Studies
• Computational Social Science

Background:

• Online social networks are rapidly evolving, making link prediction a critical research area.
• Existing link prediction methods primarily focus on network structure and node attributes.
• The influence of information diffusion on link creation and prediction remains underexplored.

Purpose of the Study:

• To investigate the impact of information diffusion on link creation and prediction within social networks.
• To propose and validate a hypothesis that information diffusion influences link formation.
• To enhance link prediction performance by incorporating features from the information diffusion process.

Main Methods:

• Data analysis of real-world user interactions on Sina Weibo, a major microblogging platform.
• Development of a novel approach that integrates information diffusion characteristics into link prediction models.
• Empirical validation of the proposed hypothesis using extensive experimental analysis.

Main Results:

• The study confirms that information diffusion patterns are significant predictors of link creation.
• A key feature derived from the information diffusion process was identified and utilized.
• Experimental results on the Sina Weibo dataset demonstrate a marked improvement in link prediction accuracy.

Conclusions:

• Information diffusion is a crucial factor to consider in understanding and predicting link formation in social networks.
• The proposed method, leveraging information diffusion features, offers a more effective approach to link prediction.
• This research provides valuable insights for social network analysis and the development of advanced prediction models.