Eur. Phys. J. B (2016) 89: 46
https://doi.org/10.1140/epjb/e2016-60509-9

Regular Article

Recommendation in evolving online networks

¹ Web Sciences Center, School of Computer Science and Engineering, University of Electronic Science and Technology of China, Chengdu 610054, P.R. China
² School of Systems Science, Beijing Normal University, Beijing 100875, P.R. China
³ Chongqing Institute of Green and Intelligent Technology, Chinese Academy of Sciences, Chongqing 400714, P.R. China

^a e-mail: an.zeng@unifr.ch

Received: 26 June 2015
Received in final form: 20 December 2015
Published online: 17 February 2016

Abstract

Recommender system is an effective tool to find the most relevant information for online users. By analyzing the historical selection records of users, recommender system predicts the most likely future links in the user-item network and accordingly constructs a personalized recommendation list for each user. So far, the recommendation process is mostly investigated in static user-item networks. In this paper, we propose a model which allows us to examine the performance of the state-of-the-art recommendation algorithms in evolving networks. We find that the recommendation accuracy in general decreases with time if the evolution of the online network fully depends on the recommendation. Interestingly, some randomness in users’ choice can significantly improve the long-term accuracy of the recommendation algorithm. When a hybrid recommendation algorithm is applied, we find that the optimal parameter gradually shifts towards the diversity-favoring recommendation algorithm, indicating that recommendation diversity is essential to keep a high long-term recommendation accuracy. Finally, we confirm our conclusions by studying the recommendation on networks with the real evolution data.