Brought to you by EarthWeb
IT Library Logo

Click Here!
Click Here!

Search the site:
 
EXPERT SEARCH -----
Programming Languages
Databases
Security
Web Services
Network Services
Middleware
Components
Operating Systems
User Interfaces
Groupware & Collaboration
Content Management
Productivity Applications
Hardware
Fun & Games

EarthWeb Direct EarthWeb Direct Fatbrain Auctions Support Source Answers

EarthWeb sites
Crossnodes
Datamation
Developer.com
DICE
EarthWeb.com
EarthWeb Direct
ERP Hub
Gamelan
GoCertify.com
HTMLGoodies
Intranet Journal
IT Knowledge
IT Library
JavaGoodies
JARS
JavaScripts.com
open source IT
RoadCoders
Y2K Info

Previous Table of Contents Next


5. THE STATE OF THE ART

Expert systems for accounting tasks are now in use throughout all parts of the accounting profession. However, individual firm or company use is still very spotty. Intelligent questionnaires that were once developed as stand-alone expert system applications are now integrated into standard auditing and tax software. Personal financial planning services once affordable by only the very well off are now being marketed to families with moderate incomes at a price they can afford. Expert system applications, including neural networks and case-based reasoning, are in use. Following are some interesting examples of expert systems at work in accounting practice today.

PaperLess Management (Brown and Phillips, 1995) is seven clusters of 25 computer programs, designed to provide individual stores, corporate headquarters, and field operations with the information necessary to manage a multilocation retail business efficiently. Many of the modules are rule-based and two, Operations Expert System and General Ledger Expert System, incorporate an inference engine. The payoff from PaperLess Management takes many forms, including the following reported by one user: reduction of management training time and the time managers spend in the hiring process, saved about 4% of the company's labor costs by reducing overtime, better production planning saved an amount equal to 6% of daily sales. PaperLess Management's systems reduce stress, thereby reducing turnover. At corporate level, the systems are used to reduce the number of people necessary to run the business.

FALCON (Brown and Phillips, 1995) is a credit card fraud prevention system used by several of the largest credit card companies. FALCON incorporates neural network, database, rule-based, and statistical modeling capabilities. It creates an individual user behavior file for each account using pools of fraud data. FALCON examines each transaction, looking for patterns that may be fraudulent. If FALCON decides that fraud is likely, it provides strategies for follow-up. According to FALCON's developer, HNC, Inc., the system has been saving clients from 20 to 50% over what their previous fraud detection systems could deliver.

Retail Sales Prediction (Brown and Phillips, 1995) is a commercially used expert system that incorporates case-based reasoning to predict daily sales volume for each individual store. Predictions are made 2 weeks in advance so they can be used by store managers to determine the number of sales people to schedule on a daily basis. The system uses similar days, including the closest same day of the week (e.g., Tuesday) and days from a holiday (e.g., Christmas eve) for a baseline. The baseline sales are then adjusted for differences in the year, special events, school holidays, etc. The system makes predictions with only about 8% variability from actual.

Planet is used by Price Waterhouse audit professionals for expert planning assistance. Planet's expert knowledge is derived from the expertise of Price Waterhouse audit partners and managers worldwide. Planet identifies risks based on auditors' responses to a series of questions and creates an audit plan that addresses those risks. With Planet's assistance, an efficient and focused audit plan can be produced in significantly less time. The audit engagement teams in several countries who first used Planet following its preliminary release in October of 1994 were so happy with it that they used it to produce their audit plans even after the test period was over. The full release of Planet began in September of 1995.

Comet, also developed by Price Waterhouse, provides support for evaluating internal controls based on modeling and analyzing business processes. Comet assists users in building a hierarchical, structured, annotated flowchart describing a business process and associated internal controls. Comet uses a proprietary approach to analyze potential errors, the relative effectiveness of compensating controls, and control weaknesses. Comet is designed for financial service clients and multinational firms, where auditors expect that high audit reliance can be placed on system controls. In addition, Comet is considered particularly useful for new clients or when a client is changing computerized information systems. Comet won a 1996 AAAI Innovative Applications of Artificial Intelligence award. Additional information on Planet, Comet, and other Price Waterhouse expert systems may be obtained on their web site, [http://www.pw.com/].

MomsExperten uses basic expert system technology to support the activities of Swedish VAT offices by dealing with the large volume of simple questions and the training of new, inexperienced staff [http://www.attar.com/pages/case_vat.htm]. With rising costs of experienced staff and increased pressure to keep costs down, even systems that use only very basic expert systems technology are likely to increase.

A very good glimpse of current research efforts in accounting expert systems can be found by looking at recent special issues on accounting applications of two journals, Expert Systems With Applications 9(4), 1995 and the International Journal of Intelligent Systems in Accounting, Finance and Management 3(3), 1994; 4(3), 1995; and 5(3), 1996. Research topics represented in these special issues include:

  • Selecting suitable accounting tasks for expert system development (1)
  • Knowledge elicitation for financial expert systems (3)
  • Cooperating expert systems (1)
  • Prototype development (3)
  • Decision bias with implications for expert systems (1)
  • Issues related to concurrent use by multiple users of the same expert system (1)
  • Integrating expert systems and other accounting modeling methods (e.g., database or REA) (2)
  • Neural networks applied to accounting tasks (3)
  • Case-based reasoning applied to accounting tasks (1)
  • Belief functions and belief networks (2)
  • Genetic algorithms (1)
  • The value and effects of explanation in accounting expert systems (2)
  • Validation of accounting expert systems (1)
  • Impacts of accounting expert systems (2)
  • Legal liability related to the use of accounting expert systems (1)
  • Diffusion of accounting expert systems (1)
  • Citation analysis (1)

Examining current research leads to the following observations on open research issues in accounting expert systems.


Previous Table of Contents Next

footer nav
Use of this site is subject certain Terms & Conditions.
Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details.