A Framework to Analyze User Interactions in an E-Commerce Environment

A Framework to Analyze User Interactions in an E-Commerce Environment

Manoj A. Thomas (Virginia Commonwealth University, USA) and Richard Redmond (Virginia Commonwealth University, USA)
DOI: 10.4018/978-1-60960-595-7.ch002


As e-commerce applications proliferate the Web, the cognitive load of sifting through the copious volumes of information in search of relevance has become formidable. Since the nature of foraging for information in digital spaces can be characterized as the interaction between internal task representation and the external problem domain, we look at how expert systems can be used to reduce the complexity of the task. In this chpater, we describe a conceptual framework to analyze user interactions in an e-commerce environment. We detail the use of the ontology language OWL to express the semantics of the representations and the use of SWRL rule language to define the rule base for contextual reasoning. We illustrate how an expert system can be used to guide users by orchestrating a cognitive fit between the task environment and the task solution.
Chapter Preview


Digital domains on the Internet exist in different forms serving different purposes. They range from e-commerce storefronts to digital libraries, offering goods and services, and serving as store houses of information that are rich in content. Visitors to these information spaces rely on navigational aids (e.g., hyperlinks, menu structures, keyword search, etc.) to find the information content of interest and to evaluate their relevance (e.g., feedback forms, reviews, ranking, testimonials, etc.). Users of these information spaces almost often have uniquely different information needs. While one might search the Amazon website (www.bizrate.com) to determine the specifications and best price of a certain brand of LED high definition TV. Ironically, it is the navigational aids that provide effective information search and enhance the consumer experience. But, they are almost always passive facilitators.

There is no shortage of information on the World Wide Web today. In our efforts to find what we need on the web, we often end up playing the guessing game of making the right choice from a finite set of juxtaposed options (links, search results, etc.). The passive navigational aids at our disposal are often not adept in leading the user in the right direction nor are they designed to reflect the cognitive process of the individual user who ultimately use them. Since browsing information spaces involve cognitive actions of the user, it is only natural to assume that navigational aids should be in agreement with, and adjusts in an anticipatory manner based on the perceived user intentions.

We undertake different types of tasks during our interaction with information spaces. We can refer to these as task environments (Newell et al., 1972). The activities of individuals in these task environments are situated in the social and physical setting in which they occur (Klahr et al., 1989; Newell et al., 1972). This applies to both physical and digital worlds, where consciously or subconsciously we engage, interact and converse with the environment before directing an action. Even today, existing information processing models fail to adequately address this situated character of activity (Suchman, 1987). For effective interaction with the environment, an individual’s knowledge structure must incorporate detailed understanding of the structural features of the environment (Klahr et al., 1989). Unfortunately, this becomes a far-fetched expectation, especially in instances where representations fail to provide any form of guidance prior to the decision to act.

Simon (1996) argues that the complexity of the environment rather than the complexity of the mechanism governs the actions of the decision maker. However, the complexity of the external environment is not the only input or stimuli that determine a person’s ability to successfully (or optimally) complete a task. The situated actions are also driven by the internal representation of the task (Khatri et al., 2006; Shaft et al., 2006; Vessey, 1991; Zhang et al., 1997). Problem solving involves human thinking that is governed by arranging simple information processes into orderly, complex sequences that are responsive to and adaptive to the task environment where clues are extracted from the environment as the sequences unfold (Newell et al., 1972; Simon, 1996). In other words, the activities of an individual in a task environment involve processing of information distributed across the internal mind as well as the external environment. The role of information perceived from external representations and the information retrieved from internal representations, and the interwoven relationship between them that leads to a decision to act is not a radical new way of thinking in this world. In this chpater we limit ourselves to the digital realm, and we use ontologies to model and discover user intentions, which in turn serves as a facilitator to guide a user through a digital information space.

Complete Chapter List

Search this Book: