Dialogue-based interaction with a web assistant: The ADVICE approach

Searching for and selecting complex products on the web is usually a difficult task for consumers mainly due to the lack of intelligent support or assistance. An advanced solution in an e-commerce setting has to optimize the user search by a customer adapted intelligent assistance that emulates in some way the performance of a human seller. This assistance should comprise customeradapted suggestion and explanation of product types, features, alternatives and special offers on digital markets. The ADVICE approach allows this enhancement of the interaction with the user by: (a) A fluid communication with the user supported by different media in order to achieve an advanced multimedia user interface. (b) A dialogue-based interaction to help user navigation in the commercial web, supporting a genuine integration of semantics and pragmatics, based on a discourse model with a twofold dimension: informational and intentional. (c) A customized configuration of the offer according to the user needs and the evolution of the dialogue. Moreover, improving system ergonomy can only be place-holder for the system intelligence. The implemented version to reach a first step on intelligent assistance in ecommerce is based on an agent-based architecture. The prototype includes the Interface Agent to manage the multimedia presentation, the Interaction Agent to support an advanced user-system interaction and the Intelligent Agent incorporating a knowledge-based model of the e-business and domain related knowledge, that supports the reasoning for advise-giving according with the user needs and the dialogue evolution. This work has been developed in the framework of the European Commission research project ADVICE (IST 1999-11305). The resulting prototype allowed the validation of the global approach for the development of advanced interfaces in web applications.