Chapter

Rough Sets and Knowledge Technology

Volume 4481 of the series Lecture Notes in Computer Science pp 109-117

Mining Associations for Interface Design

  • Timothy MaciagAffiliated withDepartment of Computer Science, University of Regina, 3737 Wascana Parkway, Regina, SK, S4S 0A2
  • , Daryl H. HeptingAffiliated withDepartment of Computer Science, University of Regina, 3737 Wascana Parkway, Regina, SK, S4S 0A2
  • , Dominik ŚlȩzakAffiliated withInfobright Inc., 218 Adelaide St. W, Toronto, ON, M5H 1W8
  • , Robert J. HildermanAffiliated withDepartment of Computer Science, University of Regina, 3737 Wascana Parkway, Regina, SK, S4S 0A2

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Consumer research has indicated that consumers use compensatory and non-compensatory decision strategies when formulating their purchasing decisions. Compensatory decision-making strategies are used when the consumer fully rationalizes their decision outcome whereas non-compensatory decision-making strategies are used when the consumer considers only that information which has most meaning to them at the time of decision. When designing online shopping support tools, incorporating these decision-making strategies with the goal of personalizing the design of the user interface may enhance the overall quality and satisfaction of the consumer’s shopping experiences. This paper presents work towards this goal. The authors describe research that refines a previously developed procedure, using techniques in cluster analysis and rough sets, to obtain consumer information needed in support of designing customizable and personalized user interface enhancements. The authors further refine their procedure by examining and evaluating techniques in traditional association mining, specifically conducting experimentation using the Eclat algorithm for use with the authors’ previous work. A summary discussing previous work in relation to the new evaluation is provided. Results are analyzed and opportunities for future work are described.

Keywords

Association mining clustering rough sets usability personalization