Affiliations: Laboratoire Heudiasyc, Université de Technologie de
Compiègne, BP 20.529; 60205, Compiègne, France. Tel.: +33 3 44 23 44 66;
Fax: +33 3 44 23 44 77; E-mail: {eparaiso,barthes}@hds.utc.fr
Note: [] Corresponding author
Abstract: This paper describes the design of an intelligent speech interface
for Personal Assistants applied to knowledge management. We believe that the
use of speech facilitates the interaction between human and agent, since the
user may speak with the agent using her own terms, increasing the quality of
the assistance. One of the difficulties is to handle spoken natural language,
understanding its actual context. To do so, we restrict the exchanges to the
following Directives Speech Act classes: inform, request, or answer. In
addition, for effective knowledge handling, we are using a set of task and
domain ontologies, separating domain and task models for reasoning. We present
such an architecture in a multi-agent system and apply it to knowledge
management. From the user's point of view, the system is a Service Centre
and the personal assistant is its Service Provider. As a result of this
conversational speech interface, we expect an increase in the quality of
assistance and a reduction of the user's cognitive load.
Keywords: Intelligent speech interface, dialogue systems, personal assistants, knowledge management
