Consumer Modelling in Support of Interface Design

Abstract

There is significant interest in developing new methods to design more effective user interfaces for decision support tools in online shopping environments. Many online companies have already begun to provide their consumers with enhanced user interface options, such as the ability to customize and/or personalize their user interface. However, for these enhanced options to produce meaningful, useful results, consumers are often required to input substantial amounts of information, placing a strain on the consumers' cognitive decision-making abilities and disrupting their focus on their immediate decision task(s). In this paper, the authors describe a personalization technique to reduce the amount of consumer information required to develop and deploy systems providing these enhanced options. Over the course of the three experiments, the authors built upon each experiment utilizing a combination of traditional statistical methods and rough set theory. This paper will describe the refined technique and the procedures, algorithms, observations, and analysis of the experiments conducted. As well, a discussion detailing future work will be provided.

Mining Associations for Interface Design

“…Consumers would be placed in the appropriate cluster based on their response and a customizable and personalized user interface would be displayed specific to the cluster group's preferences. The work described in this paper will further refine the procedure described in Maciag et al [3] (Figure 1). Specifically, the research described here will build on previous work by examining and evaluating techniques in association mining to aid in the task of designing the personalized aspects of the user interface after the initial clustering procedure, as described in Maciag et al [3], is performed.…”

“…Compensatory decision strategies are used when the decision maker applies a strict, fully rationalized thought process based on pre-defined preferences, ratings, or rankings to formulate a final decision [3]. The decision-maker will systematically weigh all possible alternatives in order to form the best possible decision outcome.…”

“…Depending on the amount of data and information that consumers must provide, this task has potential to be a highly complex and time consuming. Maciag et al [8,3] describe a technique to reduce the complexity of this task by reducing the amount of consumer information required. The primary idea of their research was to formalize the foundations of a personalization procedure aimed at clustering consumers into groups bearing similar attribute values and product preferences.…”

“…In Maciag et al [8,3], web-based shopping support tools were developed to conduct a usability evaluation. The authors chose to base their evaluation using a software support tool designed by the United States Environment Protection Agency (US-EPA) that enabled product comparisons between 29 environmentally preferable cleaning products using eight product attributes.…”

“…This information was used to develop a procedure to gather consumer information in support of clustering the participants into groups having similar attribute and product preferences. Figure 1 illustrates the authors' procedure in Maciag et al [3]. First, the 29 cleaning products were clustered, generating four product clusters.…”