Triadic co-clustering of users, issues and sentiments in political tweets
作者：
Highlights：
• Person-issue-keyword relationships models the sentiment of a person on an issue.
• On this signed 3-partite graph a new strong co-clustering problem is defined.
• A novel algorithm has been proposed and compared with spectral clustering method.
• Experiments are done on real-life data using twits of UK politicians on Brexit.
• Scalability of the new algorithm has been shown through syntactic data.
摘要
•Person-issue-keyword relationships models the sentiment of a person on an issue.•On this signed 3-partite graph a new strong co-clustering problem is defined.•A novel algorithm has been proposed and compared with spectral clustering method.•Experiments are done on real-life data using twits of UK politicians on Brexit.•Scalability of the new algorithm has been shown through syntactic data.
论文关键词：Social network analysis,Co-clustering,Hypergraph,3 partite graph,Sentiment analysis
论文评审过程：Received 7 October 2017, Revised 25 January 2018, Accepted 26 January 2018, Available online 2 February 2018, Version of Record 8 February 2018.
论文官网地址：https://doi.org/10.1016/j.eswa.2018.01.043
