Duan (2005) web-based expert systems - benefits and challenges

Information & Management 42 (2005) 799–811 www.elsevier.com/locate/dsw

Web-based expert systems: benefits and challenges
Y. Duana,*, J.S. Edwardsb, M.X. Xuc
Luton Business School, University of Luton, Luton LU1 3JU, UK Aston Business School, Aston University, Birmingham B4 7ET, UK c Portsmouth Business School, University of Portsmouth, Portsmouth PO4 8JF, UK
b a

Received 27 June 2003;received in revised form 28 May 2004; accepted 21 August 2004 Available online 8 December 2004

Abstract Convergence of technologies in the Internet and the field of expert systems have offered new ways of sharing and distributing knowledge. However, there has been a general lack of research in the area of web-based expert systems (ES). This paper addresses the issues associated with the design,development, and use of web-based ES from a standpoint of the benefits and challenges of developing and using them. The original theory and concepts in conventional ES were reviewed and a knowledge engineering framework for developing them was revisited. The study considered three web-based ES: WITS-advisor — for ebusiness strategy development, Fish-Expert — for fish disease diagnosis, and IMIS — topromote intelligent interviews. The benefits and challenges in developing and using ES are discussed by comparing them with traditional standalone systems from development and application perspectives. # 2004 Elsevier B.V. All rights reserved.
Keywords: Expert systems; Knowledge based systems; Internet; Benefits and challenges; Web-based

1. Introduction Expert systems (ES) emerged as a branch ofartificial intelligence (AI), from the effort of AI researchers to develop computer programs that could reason as humans [6]. Many organizations have leveraged this technology to increase productivity and profits through better business decisions [5,10,11,19,26]. ES are one of most commercially
* Corresponding author. Tel.: +44 1582 743134; fax: +44 1582 743172. E-mail address:yanqing.duan@luton.ac.uk (Y. Duan).

successful branches of AI [17]. Although there have been reports of ES failures [18,27], surveys [15,28] show that many companies have remained enthusiastic proponents of the technology and continue to develop important and successful applications. The early applications of expert systems were standalone, based on mainframe, AI workstations or PC platforms. Later came LAN-baseddistributed applications. Despite their commercial success, Grove [12] pointed out that several problems and limitations are associated with traditional ES applications:

0378-7206/$ – see front matter # 2004 Elsevier B.V. All rights reserved. doi:10.1016/j.im.2004.08.005

800

Y. Duan et al. / Information & Management 42 (2005) 799–811

 Knowledge bottleneck. It is difficult to acquireknowledge from different sources. Experts are often unable to express explicitly their reasoning process.  Performance brittleness. An ES is limited in its coded expertise, which relates to a narrow domain and the ES therefore performs poorly outside its boundary.  Availability. Having the expertise provided by an ES at the place and time where it is needed is a problem when limited to the use ofa stand-alone system.  Software distribution. Updating the software and interface requires many separate installation and upgrades over time. This is often beyond the competence of the users.  Communication between distributed applications. A lack of common protocols for knowledge transfer tends to discourage designs involving co-operation or dynamic information sharing. Internet-centeredinformation and communication technologies (ICT) are changing IS applications. Power [21] argued that rapid advances in Internet technologies have opened new opportunities for enhancing traditional DSS and ES. Internet technology can change the way that an ES is developed and distributed. For the first time, knowledge on any subject can directly be delivered to users through a webbased ES. Since its...