Using Neural Networks for Pattern Association for the Online Purchase of Products

Arpan Kumar Kar XLRI School of Business & Human Resources, India

Supriyo Kumar De XLRI School of Business & Human Resources, India

Abstract
Abstract: Today, a huge percentage of all the business transactions that take place in the domain of e-commerce are dominated by online shopping after the "virtual market" conceptualization of the business. This paper focuses on how pattern association rules may be obtained from the dynamic databases generated during purchases in an e-Store to maximize the profit of the marketer. In this paper, ANN has been used as a tool for generating pattern association rules during online purchases of products to aid the cross-selling of products. For getting the rules, a methodology using artificial neural networks has been adapted for usage using an extended Delta rule for initial training of the network and a hetero-associative neural network for generating and storing the associative rules. Also, a methodology has been proposed to filter out all rules which do not add economic value to the firm and then select that rule which will meet the profit maximization objective of the marketer.

Full Text Document:	Preview PDF 542Kb
Reference:	Kar, A. K., De, S. K. (2009). "Using Neural Networks for Pattern Association for the Online Purchase of Products," . Sprouts: Working Papers on Information Systems, 9(27). http://sprouts.aisnet.org/9-27
Keywords:	Pattern association, neural networks, associated purchase prediction, data mining, business intelligence, e-commerce
Item Type:	Article - Volume 9 Article 27 (2009)
Language:	English
Email:	Arpan Kumar Kar (arpan.kar@astra.xlri.ac.in) Supriyo Kumar De (skde@xlri.ac.in)

Repository Staff Only: item control page