Plausible characteristics of association rule mining algorithms for e-commerce

Abstract

Electronic commerce includes all business conduct through information and communication technology. Development of infrastructure, telecommunications, mobile technologies, the internet and social media in recent years, made a tremendous growth in business through e-commerce. Now e-commerce is a vital part of the economic development and helps in employment, FDI and GDP growth in the country. More and more companies are now on the internet and smooth the progress of transactions over the web. A large volume of data generated by these e-commerce sites which are updated very frequently. To increase the sell, customer retention and effective decision making, association rule mining play a significant role. There are number of association rule mining algorithms designed for e-commerce. A few algorithms also support the incremental and interactive association mining. In this paper, we conducted a comprehensive study of various association rule mining algorithms that support e-commerce transactions. The shortcoming of the various existing algorithms are also identified. Some plausible characteristics proposed as well for designing an efficient algorithm of e-commerce databases, which support incremental, interactive and multi-objective association rule mining.