New approaches for link prediction in temporal social networks
Nahla Mohamed Ahmed 1, 2, Ling Chen 1, 3
COMPUTER MODELLING & NEW TECHNOLOGIES 2014 18(2) 87-94
1 College of Information Engineering, Yangzhou University, Yangzhou China, 225009
2 College of Mathematical Sciences, Khartoum University, Khartoum, Sudan, 115547
3 State Key Lab of Novel Software Tech, Nanjing University, Nanjing China, 210093
Link prediction in social networks has attracted increasing attention from various domains such as sociology, anthropology, information science, and computer sciences. In this work, efficient approaches to predict potential link in temporal social networks are presented. One approach is based on reduced static graph using a modified reduced adjacency matrix to reflect the frequency of each link. Another approach is based on indices integration，and exploits both the temporal and topological information. The approach integrates the indices in all the time steps, which reflect the topological information, and uses a damping factor to emphasize the importance of more recent links. Experimental results on real datasets show that our approaches can efficiently predict future links in temporal social networks, and can achieve higher quality results than traditional methods.