Data mining for mobile app users identify based on improved RFM model in a case of social network

Despite the high demand and the accelerated growth of mobile apps during the last years, only a few studies have been done to the multi-dimensional evaluation of the app users. In this paper, the RFM General Model refers to the critical gap in the literature and illustrates how the users of mobile apps behave in terms of novelty, frequency, and financial from both the messaging and financial perspective. Clustering and ranking of mobile app users have been developed in the social networking market for the first time. This study examines a wide range of customers with different characteristics in the same cluster categories and then ranks them with a simple weighting method. The most important results of the research are the three clusters of customers, with only 814 customers, that consists only 0.04% of the total customers, and in fact are the best and most profitable customers for the company. They involve 2% of total transactions and 5% of total messaging. The second group, which consists 20.6% of the total customers and involve 10% and 15% of total transactions and messaging, may lead to competing transactions. Finally, the third group, with 76% of the total customers, has a moderate downward behavior of every two perspectives, and the weakness of the users of this cluster is clearly evident. The most important usages of this article are the explanation of specific marketing policies for each cluster.

Data mining for mobile app users identify based on improved RFM model in a case of social network Keywords:

App users , RFML analysis , Data mining , Customer segmentation , SAW ranking