[an error occurred while processing this directive]

Book Contents

Ch. 10
Building Knowledge-Driven DSS and Mining Data

Chapter Contents
Previous Page
Next Page

Knowledge-Driven DSS Examples

Two classic examples of business expert systems are TAXAdvisor and XCON. More recent examples include a Scheduling System for the Tomakomai Paper Mill, a Customer Support System at Compaq Computer, and an Insurance Plan Selection System for Meiji Mutual Life Insurance Company.

TAXADVISOR was a Knowledge-Driven DSS designed to assist an attorney with tax and estate planning for clients with large estates (greater than $175,000). The system collected client data and inferred actions the clients need to take to settle their financial profile, including insurance purchases, retirement actions, transfer of wealth, and modifications to gift and will provisions. TAXADVISOR used knowledge about estate planning based on attorneys' experiences and strategies as well as more generally accepted knowledge from textbooks. The system used a rule-based knowledge representation scheme controlled by backward chaining. TAXADVISOR was implemented in EMYCIN. It was developed at the University of Illinois, Champaign-Urbana, as a PhD dissertation and it reached the stage of a research prototype.

XCON (eXpert CONfigurer of VAX 11/780 computer systems) was developed to configure computer systems. Based upon a customer's order it decided what components needed to be included to produce a complete operational system and it determined the spatial relationships among all of the components. XCON was implemented in an expert system shell called OPS5 and was developed through a collaboration between researchers at Carnegie-Mellon University and Digital Equipment Corporation (now Compaq). This commercial expert system configured VAX computers on a daily basis and was for many years the largest and most mature rule-based expert system in operation. XCON was not actually a DSS because it made decisions rather than supporting managerial decision-making.

Scheduling systems and control systems are needed in the paper production industry to ensure that all plants in the mill operate correctly. At Tomakomai Mill an expert system is used to schedule the paper production machines. The Tomakomai Mill consists of ten paper making machines, energy supply plants and pulp supply plants. Two hundred paper products are produced per month. Each product has a specified production volume and due date, and requires a specified machine to produce. In the Tomakomai Mill, a millwide production management system exists. The system has a planning level and a control/operation level. The scheduling system is situated on the planning level for papermaking. It receives product orders from the headquarters office, makes a schedule and delivers it to the other planning systems. Each system schedules and optimizes its operations based on the papermaking schedule. The paper production scheduling system consists of an expert system for automated scheduling, and a data management system. The expert system consists of three subsystems: product group scheduling system, individual scheduling system, and a balancing scheduling system. The scheduling systems are implemented with an expert shell utility, ASIREX. This scheduling system has been in practical use since January 1989. The greatest advantage of this system is that it speeds up scheduling. The scheduling time for a monthly schedule was reduced from 3 days to 2 hours.

Compaq Computer Corporation created and implemented a very successful Customer Support Intelligent System to provide computer users expert diagnosis and recommendations about problems with hardware, software, network and other problems (cf., Dhar and Stein, 1997).

The Meiji Mutual Life Insurance Company is one of the oldest life insurance companies in Japan with assets of around $74 billion. Meiji offers a wide range of insurance and pension products. In addition, the company is aggressively involved in developing and introducing new products. However, with the increasing number of products, the company was finding it difficult to ensure that all the insurance sales staff had the expertise and the latest knowledge required to provide the best advice and service to its customers. To overcome this problem, Meiji used XpertRule to develop the Life Insurance Plan Selection Expert System. The system can select the most suitable product, along with a reason for the choice, from Meiji's range of 37 individual oriented products. Meiji began research into expert systems in 1986. Before using XpertRule, the company had completed a Lisp-based insurance plan selection system. This system, however, had a high delivery and maintenance cost and was not suited for distribution to all branches. Meiji adopted XpertRule because it allows for easy knowledge base construction. The knowledge base contained 47 decision tasks. The rules for selecting each plan were developed as a separate task. The system was structured so that when the details of a customer are entered, the system assesses the suitability of all the plans and report on the best five. The system only takes 3 to 4 seconds to make suitable selections (cf., http://www.attar.co.uk/pages/case_ml.htm).

[an error occurred while processing this directive]