Deliberative Stock Market Agents using Jinni and Defeasible Logic Programming
Alejandro Javier García, Devender Gollapally, Paul Tarau, Guillermo Ricardo Simari · 2000
Working with stock markets requires constant monitoring of the stock information which keeps changing continously, and the ability to take decisions instantaneously based on certain rules as the changes occur. In this paper, a framework for implementing a Deliberative MultiAgent System is developed. This system can be used as a proactive tool for expressing and putting to work high level stock trading strategies. In this framework, agents are able to monitor and extract stock market information via the World Wide Web and, using the domain knowledge provided in the form of defeasible rules, can reason in order to achieve the established goals. The overall system is integrated using Jinni which provides a platform for building intelligent autonomous agents. The agents have a Reasoning Module, based on Defeasible Logic Programming, capable of formulating arguments and counterarguments in order to decide whether or not to perform an action. Arguments and counterarguments are compared and a thorough dialectical analysis is performed. From the outcome of this analysis agents are able to decide which action to perform based on the support of warranted arguments, as real world agents would do.