Purchasing Pattern of the Customers Based on Outliers Analysis

Abstract

Data Mining (referred as extracting knowledge from data) is the process of discovering patterns, associations, and links in the huge stack of data which is based on the analysis done through different perspectives. There are many disciplines which are found under data mining some of them are clustering analysis, regression analysis, and classification analysis and so on. To analyse the data, we need some sort of software which can be used to perform the required analysis on data set. In this paper, the researchers are using WEKA library (.jar file) for data mining and implementing the outlier detection algorithm for finding the outliers from the data set of a European client. The outliers are the distinct points in the data which are extremely different from the rest of the data set which can lead to inappropriate outcomes. These outlier observations can deviate so much from other observations, as to arouse suspicion that it was generated by a different mechanism. The researchers are trying to find out the distinct purchasing behavior of the customers and categorize them based on their different purchasing trends like improving, leading, challenging and lagging and then the outlier and exception customers are identified who lies outside the quadrant region. This study helps e-commerce merchants and vendors to identify the customers who are giving them less sale and those who are giving them more sales in order to handle them accordingly by giving promotional offers or discounts.