A User Influence Network Construction Approach Based on Web Mining and Social Network Analysis.
Wenyu Yuan,Zhen Zhang,Danni Chang
DOI: https://doi.org/10.1109/IEEM58616.2023.10406593
2023-01-01
Abstract:With the rapid development of Mobile Internet, user-generated contents have become important user data sources. It has been well recognized that users' attitude and purchase decisions may be influenced by others. However, it has not been fully revealed that how such influence will be formed, and how users' attitudes and requirements will be changing under the user influence. To tackle the problems, a data-driven approach was proposed to track the user influence on their product/service requirements, and then construct a user influence network to show the specific connections. To illustrate the proposed approach, a case study on RED (a social media APP) was performed. More than 500 fans of electric car were investigated. Based on their online behaviors, such viewing, reposting, or liking, a time-series user social network was constructed. The results show that the main EVs features that attract users remain similar, but the specific focus, such as the fashion color, grade, is changing under user influence. It can help enterprises to deeply understand the essential product features from user perspective, and the dynamic user expectations.