ISBN: 978-981-18-3959-7 DOI: 10.18178/wcse.2022.06.013
Research on Early Warning of User Churn in Social Platform based on Data Mining Technology
Abstract— The management of user churn is essential content in the enterprise customer management of the social platform. By constructing the early warning model of user churn, we can predict the potential lost users so that enterprises can give early warning and take corresponding measures to retain users and reduce the cost of maintaining users, which has specific practical significance.
Research the application of data mining technology in predicting user churn in the social media industry. Using data mining technologies such as Logistic regression algorithm, Bayesian algorithm, and XGBoost algorithm, the churn model is constructed, and the models are compared and integrated so that the model with high prediction accuracy can be obtained. The potential churn users can be predicted, which is the key user group maintained by enterprises.
Index Terms—social media users, Data mining, Loss model, voting fusion , Churn prediction
Wangyang Shi
Technological University of the Philippines Manila, PHILIPPINES
Anhui Technical College of Mechanical and Electrical Engineering Wuhu, CHINA
Jonathan M. Caballero, Jonan Rose Montana
Technological University of the Philippines Manila, PHILIPPINES
Cite: Wangyang Shi, Jonathan M. Caballero, Jonan Rose Montana, "Research on Early Warning of User Churn in Social Platform based on Data Mining Technology, " Proceedings of 2022 the 12th International Workshop on Computer Science and Engineering (WCSE 2022), pp. 80-85, June 24-27, 2022.