Data Mining for Needy Students Identify Based on Improved RFM Model: A Case Study of University

Abstract

The government has built up a set of support system for poverty-stricken students in colleges and universities. There is a contradiction between the shortages of higher education tuition fees with the popularization of higher education. Limited funds allocated require resources and targeted toward needy students. RFM (Recency, Frequency, and Monetary) method is very effective attributes for customer segmentation. Our goal in this paper is to build improved RFM-based customer segmentation model to identify those needy students through the database of dining room. This study first build a framework for identify needy students based on RFM. Then improved RFM model through redefine Recency as ratio to data mining, and used Analytic Hierarchy Process (AHP) to determine weights of RFM variables. Finally, this study applied K-means algorithm to identify needy students. Through case study, the method can efficiently identify needy students and provide needy students list to department of university as reference.