Using agent technology for exception monitoring -The case of Fraud detection in conference management systems
Abstract
Exception monitoring involves solutions to overcome the fatality of exceptions. The general idea behind this concept is that exceptions should be monitored, and as soon as they occur, they should be eliminated in one way or another so that the systems keep on working without being impacted by the potential harm of these unusual situations present. Misuse or fraud is one of such exceptional situations. In this thesis, a monitoring agent application, MONA, is designed, developed and evaluated in order to support business processes for capturing fraudulent activities with an empirical focus on the case of conference management systems. With the help of this prototype, also the time and quality impact of agent technology, and if agent technology creates an increase in time effectiveness and enhances the quality of work, are investigated. Another focal point of the project is to find out if the results of collaborating with the monitoring agent are reliable, and whether they are convincing enough for further collaboration. In the evaluation process usability test is conducted through observations and interviews with the potential users. The results indicate that there are gains in time, time-effectiveness increases, and the quality of fraud detection work is enhanced. Another important result is that the test participants are likely to trust MONA.