Supporting online shopping through a combination of ontologies and interface metaphors

In this paper we describe some results of the Alice project. Alice is an ontology-based e-commerce project which aims to support online users in the task of shopping. Ontologies describing customers, products, typical shopping tasks and the external context form the basis for the Alice architecture. We also build upon two novel interface metaphors originally developed for navigating databases: the Guides metaphor and Dynamic Queries. The Guides metaphor was developed at Apple to reduce the cognitive load on learners navigating a large hypermedia database. Within Alice we use the Guides metaphor to allow online shoppers to interact with the system in a variety of ways. In effect, by choosing these options they are classifying themselves for the purpose of customizing system responses. We discuss the link between Alice Guides and Kozinets’ notion of e-tribes or Virtual Communities of Consumption. Our second interface metaphor Dynamic Queries (coupled with Starfield displays) allows users to very quickly find relevant items by displaying the results of queries, posed via specialized slider widgets, within 100 ms. We have constructed a tool, Quiver, which constructs Dynamic Query interfaces on-the-fly as the result of queries to knowledge models stored on the Alice server.