Title: Multi-feature fusion friend recommendation algorithm based on complex network

Authors: Kan Pan; Hailong Chen; Qian Liu; Jian Wang; Yingming Pu; Chunlin Yin; Zheng Yang; Na Zhao

Addresses: Electric Power Research Institute, Yunnan Power Grid Co., Ltd., Kunming 650217, China ' School of Software, Yunnan University, Kunming 650504, China ' School of Software, Yunnan University, Kunming 650504, China ' College of Information Engineering and Automation, Kunming University of Science and Technology, Kunming 650504, China ' School of Software, Yunnan University, Kunming 650504, China ' Electric Power Research Institute, Yunnan Power Grid Co., Ltd., Kunming 650217, China ' Electric Power Research Institute, Yunnan Power Grid Co., Ltd., Kunming 650217, China ' Electric Power Research Institute, Yunnan Power Grid Co., Ltd., Kunming 650217, China; School of Software, Yunnan University, Kunming 650504, China

Abstract: At present, one of the problems of friend recommendation algorithms used in most social networks is that these networks often rely on a single index for recommendation. To solve this problem, multi-feature fusion (MFF) algorithm, a social network friend recommendation algorithm based on complex network theory, is proposed. The recommendation algorithm works by firstly divides the existing social networks into different communities. The importance of nodes in a social network is then calculated through the fusion of nodes' importance information. Lastly, by integrating node importance information, friend number information and the shortest path information features are comprehensively evaluated, so as to generate final friend recommendation list. Simulation shows that with the increase of network nodes, the MFF algorithm outperforms common friend (CF) algorithm and friend similarity (FS) algorithm over all evaluation indicators including P-value, R-value and F1-value.

Keywords: complex network; social network; friend recommendation; node importance; multi-feature.

DOI: 10.1504/IJICT.2023.134831

International Journal of Information and Communication Technology, 2023 Vol.23 No.4, pp.401 - 423

Received: 28 Sep 2021
Accepted: 25 Oct 2021
Published online: 14 Nov 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article