LRFMP model and K-means clustering for customer profiling in online retail market.

Customer profiling is one of the important aspects for retailers or service providers used in analyzing consumer characteristics and needs. To build a customer profile, various customer information can be used ranged from demographic information (such as age, gender, salary, address) to customer behavior (such as the frequency to visit shop and money spent). Due to the enormous amount of information available, it becomes inefficient to analyze it through descriptive analysis. Instead, data mining can be used to cluster and summarize the huge amount of information without much information loss. This paper discuss method of utilizing LRFMP (Length Recency Frequency Monetary Periodicity) model and K-means clustering in data mining techniques to perform customer profiling. Furthermore, this paper will describe the evaluation of this work which is based on Elbow method, Silhouette Index, Davies-Bouldin Index, and Calinski Harabasz test. The result suggested LRFMP model is a viable model in customer profiling when use together with K-means clustering technique despite some variable may had contributed less than another. [ABSTRACT FROM AUTHOR]