Affiliations: [a] School of Design, Jiangnan University, Wuxi, China
| [b] NanHang Jincheng College, Nanjing, China
| [c] Jiangsu Key Laboratory of Media Design and Software Technology, Jiangnan University, Wuxi, China
Correspondence:
[*]
Corresponding author. Yuan Liu, Jiangsu Key Laboratory of Media Design and Software Technology, Jiangnan University, Wuxi, 214122, China. Tel.: +86 0510 85327305; E-mail: lyuan1800@sina.com.
Abstract: As a communication tool, social networks have become an important part of people’s daily lives. The proliferation of data in the big data era has spurred waves of research. Twitter is one of the most popular sites among all social network platforms. Twitter API allows researchers to easily study user behaviour and influence. In this paper, we use data obtained through the Twitter API to study the influence maximization problem based on the relationship graph of the social network and information dissemination model. In the current independent cascade propagation model, the influence weights between nodes are based largely on the fixed-value assumption and the randomized probability of acceptance; however, this does not conform to real life. The influence weight between users is closely related to the strength of relationships, content of communication, and so on. Focusing on user relation-ships, we introduce an improved weighted cascade mode combined with a heuristic algorithm to find an approximate solution to the problem of influence maximization. Example analysis indicates that the improved weighted cascade model can obtain a more significant and influential node set compared to conventional methods.
Keywords: Social network, user behaviour, user Influence
