Friends recommendation on social networks

• Main Author: Rahman Syukri Othman
• Other Authors: Mahardhika Pratama
• Format: Final Year Project
• Language: English
• Published: 2019
• Subjects: DRNTU::Engineering::Computer science and engineering
• Online Access: http://hdl.handle.net/10356/76993
• Institution: Nanyang Technological University

Social networks like Facebook and Twitter are prevailing nowadays. People use social networks to stay up-to-date with news and current events. Besides that, most people use social network so that they can make new friends. Friends recommendation on social networks is vital as it allow people to build better relationships as well as promoting information sharing and spreading. This project will investigate the friend recommendation problem for social networks. Unlike existing techniques like collaborative filtering that are widely utilized in recommender systems, this project will investigate the recommendation problem from the perspective of network science. To be specific, a social network will first be modeled as a temporal graph, then link prediction technique and its variants will be explored for calculating the probability for any pair of unconnected nodes to be connected. The recommendation methods include using evolutionary computations like Discrete Particle Swarm Optimization (DPSO) and Genetic Algorithm. Experiments on real-world social networks will be carried out to validate the effectiveness of the recommendation methods.