Churn analysis of online social network users using data mining techniques

A churn is defined as the loss of a user in an online social network (OSN). Detecting and analyzing user churn at an early stage helps to provide timely delivery of retention solutions (e.g., interventions, customized services, and better user interfaces) that are useful for preventing users from churning. In this paper we develop a prediction model based on a clustering scheme to analyze the potential churn of users. In the experiment, we test our approach on a real-name OSN which contains data from 77, 448 users. A set of 24 attributes is extracted from the data. A decision tree classifier is used to predict churn and non-churn users of the future month. In addition, k-means algorithm is employed to cluster the actual churn users into different groups with different online social networking behaviors. Results show that the churn and non-churn prediction accuracies of ∼65% and ∼77% are achieved respectively. Furthermore, the actual churn users are grouped into Ave clusters with distinguished OSN activities and some suggestions of retaining these users are provided.