Journal of Jishou University(Natural Sciences Edition)

• Computer • Previous Articles     Next Articles

Core-User-Group-Based Unbalanced Heat Conduction and  Mass Diffusion Algorithm

YIN Pengfei, WANG Guojun   

  1. (1. School of Information Science and Engineering, Central South University, Changsha 410000, China; 2. School of Information Science and Engineering, Jishou University, Jishou 416000, Hunan China; 3. School of Computer Science, Guangzhou University, Guangzhou 510006, China)
  • Online:2019-11-25 Published:2019-12-16


Mass diffusion and heat conduction algorithms are commonly used as two classic network structure-based algorithms for information filtering. However, the existing algorithms treat all users equally without distinguishing the contributions of different users and removing malicious users, which results in poor performance. In response to the above problems, we propose a core-user-group-based unbalanced heat conduction and mass diffusion algorithm (core-UHM). The underlying idea is mainly based on the Pareto Principle in the information field, that is, 20% of users dominate 80% of the interaction in the whole system. Based on core users, not only can the quality of resources be guaranteed, but also the computational overhead is greatly saved. Experiments on real academic search log data sets demonstrate the effectiveness of our proposed algorithm.

Key words: mass diffusion, heat conduction, bipartite network, information filtering, intelligent search

WeChat e-book chaoxing Mobile QQ