Effective contact recommendation in social networks by adaptation of information retrieval models

作者：

Highlights：

• IR models can be adapted as standalone algorithms to effectively recommend contacts in social networks.

• IR models are shown to be even more effective as neighbor selection methods in kNN.

• We achieve further effectiveness enhancements by learning to rank upon IR models.

• We test the researched approaches extensively in five network samples of different kind from Twitter and Facebook.

摘要

•IR models can be adapted as standalone algorithms to effectively recommend contacts in social networks.•IR models are shown to be even more effective as neighbor selection methods in kNN.•We achieve further effectiveness enhancements by learning to rank upon IR models.•We test the researched approaches extensively in five network samples of different kind from Twitter and Facebook.

论文关键词：Social networks,Contact recommendation,Information retrieval models,k nearest neighbors,Learning to rank,Collaborative filtering

论文评审过程：Received 31 October 2019, Revised 25 April 2020, Accepted 26 April 2020, Available online 19 May 2020, Version of Record 19 May 2020.

论文官网地址：https://doi.org/10.1016/j.ipm.2020.102285