The query answering system realizes the selection of the data, preparation, pattern discovering, and pattern development processes in an agent-based structure within the multi agent system, and it is designed to ensure communication between agents and an effective operation of agents within the multi agent system. The system is suggested in a way to process and evaluate fuzzy incomplete information by the use of fuzzy SQL query method. The modelled system gains the intelligent feature, thanks to the fuzzy approach and makes predictions about the future with the learning processing approach. The operation mechanism of the system is a process in which the agents within the multi agent system filter and evaluate both the knowledge in databases and the knowledge received externally by the agents, considering certain criteria. The system uses two types of knowledge. The first one is the data existing in agent databases within the system and the latter is the data agents received from the outer world and not included in the evaluation criteria. Upon receiving data from the outer world, the agent primarily evaluates it in knowledgebase, and then evaluates it to be used in rule base and finally employs a certain evaluation process to rule bases in order to store the knowledge in task base. Meanwhile, the agent also completes the learning process. This paper presents an intelligent query answering mechanism, a process in which the agents within the multi-agent system filter and evaluate both the knowledge in databases and the knowledge received externally by the agents. The following sections include some necessary literature review and the query answering approach Then follow the future trends and the conclusion.
