Scalable and Explainable Friend Recommendation in Campus Social Network System
In the recent years, social networks have been growing in popularity and importance and to a certain degree contributing to a change in human life style. In a social network system, it’s essential to offer well-designed and effective friend recommendation
- PDF / 546,508 Bytes
- 10 Pages / 439.37 x 666.142 pts Page_size
- 56 Downloads / 207 Views
Scalable and Explainable Friend Recommendation in Campus Social Network System Zhao Du, Lantao Hu, Xiaolong Fu and Yongqi Liu
Abstract In the recent years, social networks have been growing in popularity and importance and to a certain degree contributing to a change in human life style. In a social network system, it’s essential to offer well-designed and effective friend recommendation service for achieving high loyalty of users. Although Friend-Of-a-Friend (FOF) is widely used and proved to be effective, the straightforward implementation of FOF needs large amount of computation power which is a heavy burden for lightweight social network taking into account the restriction of resources. We propose a FOF-based friend recommendation algorithm in a campus social network system which is explainable and efficient. On one hand, we take multiple relationship factors into account for recommendation. On the other hand, we use incremental relationship data instead of the entire relationship data to generate latest recommendation list and detailed explanations. Ultimately, it achieves better performance in complexity and scalability. Keywords Friend recommendation and explainable
Campus social network FOF Scalable
45.1 Introduction In the recent years, social networks have been growing in popularity and importance and to a certain degree contributing to a change in human life style. Internet users tend to spend much more time on social networks than any other Internet
Z. Du (&) X. Fu Information Technology Center, Tsinghua University, Beijing 100084, China e-mail: [email protected] L. Hu Y. Liu Department of Computer Science and Technology, Tsinghua University, 100084 Beijing, China
S. Li et al. (eds.), Frontier and Future Development of Information Technology in Medicine and Education, Lecture Notes in Electrical Engineering 269, DOI: 10.1007/978-94-007-7618-0_45, Springer Science+Business Media Dordrecht 2014
457
458
Z. Du et al.
applications. Social networks become an indispensable part of people’s lives. People exchange information, express ideas and make friends in social networks. Following this tendency, many organizations like enterprises and universities attempt to construct enterprise-level or campus-level social network services to their members. For instance, IBM started to provide its enterprise social network service from September 2007 [1]. It’s valuable and essential to offer well-designed and effective friend recommendation service for achieving high loyalty of users in a social network. Friend recommendation proposes a limited, ordered and personalized list of potential friends to target users. Sometimes the recommendations are also accompanied with proper explanations which increase the probability of acceptance by users and ultimately help to enrich the connections in social network. The recommender systems can be traced back to cognitive science, approximation theory, information retrieval, forecasting theories, management science and consumer choice modeling in marketing. I
Data Loading...
