Customer Data Clustering using Data Mining Technique

TL;DR

This paper discusses using data mining, specifically customer clustering, to analyze large customer datasets for identifying high-value, low-risk customers to support business decision-making.

Contribution

It applies demographic clustering with IBM I-Miner to identify profitable customer segments, demonstrating the effectiveness of data mining in customer analysis.

Findings

Clusters represent 10-20% of customers generating 80% of revenue

Data cleansing and profiling improve clustering accuracy

Customer segmentation aids targeted marketing strategies

Abstract

Classification and patterns extraction from customer data is very important for business support and decision making. Timely identification of newly emerging trends is very important in business process. Large companies are having huge volume of data but starving for knowledge. To overcome the organization current issue, the new breed of technique is required that has intelligence and capability to solve the knowledge scarcity and the technique is called Data mining. The objectives of this paper are to identify the high-profit, high-value and low-risk customers by one of the data mining technique - customer clustering. In the first phase, cleansing the data and developed the patterns via demographic clustering algorithm using IBM I-Miner. In the second phase, profiling the data, develop the clusters and identify the high-value low-risk customers. This cluster typically represents the 10-20 percent of customers which yields 80% of the revenue.