Discovering strategic multi-agent behavior in a robotic soccer domain

2. THE MASM ALGORITHM An input to the MASM algorithm is a time-annotated multi-agent action sequence. The action sequence is then transformed into an action graph. An action graph is a directed graph, where nodes represent agent actions. Nodes a and b are connected, a → b, if action b occurs directly after action a. The position of each node corresponds to the agent position when performing action represented by the node. The process of constructing an action graph from a multi-agent action sequence is presented in work from Bezek [2.]. This paper presents a method for multi-agent strategic modeling (MASM) applied in a robotic soccer domain. The method transforms multi-agent action sequences into a visual graph-based diagram, called action graph. Graph nodes are further augmented with additional domain knowledge. Using hierarchical clustering, action graph nodes are merged by utilizing domain-specific function. This step results in an abstract graphical model of agent behavior. Then,...

• Type: Conference
• Year: 2005
• Where: ATAL
• Authors: Andraz Bezek
• Keywords: Action Graph, Action Graph Nodes, ATAL 2005, Multi-agent Action Sequences