A PROACTIVE IDENTIFICATION SYSTEM OF USER NEEDS FOR WEB SERVICES

This paper presents the implementation and a realistic case study ofMPSWS system. MPSWS is a multi-agent, context-aware and OWLS-based system that aims to support user-to-SWS interactions. In MPSWS, both user information and SWS are interacting to identify user needs. MPSWS identifies user needs through two approaches: (1) Committing approach using BDI and (2) Learning approach using AQ. In the BDI approach, MPSWS uses user desires and his contexts as inputs in the need identification process. When the user desires are not available, the learning approach is employed. In this case, MPSWS has no guidance (inputs) from the user but it uses the user history events in its need identification process. Learning of the user need is coming through AQ-based learning process. For each identified user intention, MPSWS building up OWLS queries, which are processed by SWS matchmaker, to identify OWLS services that can semantically match user needs. The presented case study shows how MPSWS proactively identifying users intentions and its semantically matched SWS depending on their current context. MPSWS is experimentally tested and its performance at different operational conditions is evaluated.