Farzad Peyravi - Electrical and Computer Engineering DepartmentYazd University, Yazd, Iran
Alimohammad Latif - Electrical and Computer Engineering DepartmentYazd University, Yazd, Iran

Today, because of explosion of non-English documents, Cross Language Information Retrieval (CLIR) has become more important. On the other hand, multi agentsystems are widely used to perform the tasks in parallel mode to handle the complexity and distribution feature of CLIR. In this paper, our designed Agent based CLIR system leads in defining four different classes of agents co-operating with each other based on MAS-CommonKADS methodology. To verify the correctness of co-operation and interaction among agents defined in a multi agent environment we modeled the system and verified it by NUSMV model checker. This is accomplished by examining agent conversations before deploying the system. Our experiments validate some of the important properties identified in this model. Results prove the agent’s specification and indicate that our proposed model works accurately

component; verification; NUSMV; crosslanguage information retrieval;