.png){kind=icon}
.png){kind=icon}
Return
Improved artificial bee colony algorithm based on community detection for link prediction problem
.png){kind=icon}
PRE
AI
DOI：10.1007/s11042-023-17197-6.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The problem of link prediction has recently gained a lot of attention from various domains, including sociology, anthropology, information science, and computer science. In this research, we formulated the link prediction problem as an optimization problem to predict the links of any type of network and proposed two artificial bee colony-based link prediction algorithms enhanced with some neighborhood structures borrowed from genetic algorithms. In an attempt to complete the network's missing links, the first algorithm optimizes weights, which are used in a linear combination with local and global similarity indexes, to capture the best of both schemes and compensate for each approach's weaknesses. In the second one, we make use of the community structure of the network to help predict the missing links present in the networks. Experimental results of the proposed algorithms on a number of real-world and synthetic networks provided interesting insights about the problem and showed that the proposed algorithms can enhance link prediction accuracy and are quite competitive with state-of-the-art algorithms.
Keywords:
Social networks
Link prediction problem
Community detection
Metaheuristics
Artificial bee colony
Journal
IF:
3
Papers:
1.9W
Citations:
3.2W