A Novel Approach for Classification of E-Commerce Data

Abstract

Classification is one of the most interesting problems in the fast evolving fields such as e-commerce and web-based businesses where the data is growing exponentially. Existing classification techniques over e-commerce data are mainly based on the users’ purchasing patterns. However, gender preferences significantly improve in recommending various products, targeting customers for branding products, providing customized suggestions to the users etc. In this paper, we propose a two-phase approach for gender based classification to classify e-commerce data by exploiting hierarchical relationships among products. The first phase reduces the dimensionality of the data by identifying the features that well describes the browsing pattern of the users. The second phase classifies the data based on these features. Experiments are carried out on clickstream data (provided by FPT group) consisting of browsing logs (with list of products formed as a hierarchy), session start time and session end time. We compared our results with standard Bayesian classification model, which shows the applicability of our classification approach for ecommerce data.