Discovering Dedicators with Topic-Based Semantic Social Networks


  • Jiyeon Jang Korea Advanced Institute of Science and Technology (KAIST)
  • Sung-Hyon Myaeng Korea Advanced Institute of Science and Technology (KAIST)



Dedication Analysis, Twitter, Topic-based Social Network, Conversation-based Social Network


Influential people are known to play a key role in diffusing information in a social network. When measuring influence in a social network, most studies have focused on the use of the graph topology representing a network. As a result, popular or famous people tend to be identified as influencers. While they have a potential to influence people with the network connections by propagating information to their friends or followers, it is not clear whether they can indeed serve as an influencer as expected, especially for specific topic areas. In this paper, we introduce the notion of dedicators, which measures the extent to which a user has dedicated to transmit information in selected topic areas to the people in their egocentric networks. To detect topic-based dedicators, we propose a measure that combines both community-level and individual-level factors, which are related to the volume and the engagement level of their conversations and the degree of focus on specific topics. Having analyzed a Twitter conversation data set, we show that dedicators are not co-related with topology-based influencers; users with high in-degree influence tend to have a low dedication level while top dedicators tend to have richer conversations with others, taking advantage of smaller and manageable social networks.




How to Cite

Jang, J., & Myaeng, S.-H. (2021). Discovering Dedicators with Topic-Based Semantic Social Networks. Proceedings of the International AAAI Conference on Web and Social Media, 7(1), 254-262.