Towards a Knowledge-Based Approach for Autonomouse Trading Agent

Abstract

The trading agent competition for supply chain manage-ment (TAC SCM) is an annual competition. An agent is a software program that acts autonomously among other agents. An agent’s goal is to maximize profit while partici-pating in a supply chain scenario. In this article we present a knowledge-based approach describing the accumulation and optimization of a knowledge base for supply chain management. A framework for designing strategies for sup-ply chain management trading agent is implemented. The strategy plan is described in simple high level language as a sequence of statements called strategy scripting lan-guage. The agent’s knowledge base and strategy builder tool is represented in this paper. Game flowchart describ-ing game and knowledge base interactions is represented. In addition we discuss the SOARS Beer Game which is a modification of the original Beer Game by Northwester University. We suggest the application of our proposed strategy design platform to this supply chain game. We have chosen TAC SCM game as a simulation platform for applying and proving our methodology. In addition we show that it can be employed by other various simulation platforms.