Title: A novel method for the selection of expert systems using case-based reasoning

Authors: Babita Pandey, R.B. Mishra

Addresses: Department of Computer Engineering, IT, BHU, Varanasi 221005, India. ' Department of Computer Engineering, IT, BHU, Varanasi 221005, India

Abstract: The selection of expert system shell (ES shell) and web-based expert system (WBES) is an important issue in their design and development. Very few literatures depict the methods for selection of stand alone or WBES. We have developed a CBR approach which considers some processes of CBR such as knowledge representation, acquisition, retrieval, learning and adaptation. A table is prepared that show the components of knowledge based system (KBS) in ES shell; KBS components and web services and languages in WBES. A simple query (binary) is used to gather and store the user requirements in the form of user data vector (UDV). ANN and heuristic methods are deployed forlearning and adaptation of a new case. The matched results show the confidence in selecting a particular ES shell and WBES and their corresponding component. The shells with maximum and near by unity confidence are selected.

Keywords: expert system selection; web services; case-based reasoning; CBR; web-based expert systems; internet; knowledge representation; knowledge acquisition; knowledge retrieval; knowledge learning; knowledge adaptation; knowledge based systems; KBS; ANNs; neural networks.

DOI: 10.1504/IJKESDP.2009.022719

International Journal of Knowledge Engineering and Soft Data Paradigms, 2009 Vol.1 No.2, pp.112 - 138

Published online: 25 Jan 2009 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article