Abstract
With the explosion growth of available information in texts, reading is a very time consuming task. Having an automated knowledge miner that can extract useful knowledge from the texts is very desirable. In this paper, we propose an integrated architecture for an intelligent knowledge mining processor. The architecture is based on text mining and data mining approaches, and a multi-agent system. In the text mining approach, we focus on extracting entity and concepts from the texts by using information extraction and natural language processing techniques. Of the data mining approaches, a classification and visualization techniques are applied to classify the extracted entities/concepts. The proposed architecture is designed using agent systems. Each process in the architecture is presented as an agent. Interactions between agents are conducted through messages. In this framework, FIPA ACL is proposed as an agent communication language. The main reason of using an agent technology is to decompose complex problem into sub problems and each problem is resolved independently.