Im Intelligent Search Robot System

Instant message and information retrieval is one of the most important Internet applications, which has a wide range of user groups and the active user's participation. There is a great number of information on Internet with large varieties and random distribution. Only depending on information retrieval we will be able to locate information on Internet accurately and rapidly and only through information retrieval we can publish information on the Internet rapidly. Information retrieval-based search rankings and search advertising business have been apart of our life.Instant message is a convenient way of communication. People can chat online, exchange documents, and make voice and video conference through it. Recently, it has been extended to support on-line customer service, map queries, etc. Instant message has entered into modern business activities.Two applications have not been well combined for a long time, which makes only text and multimedia information can be retrieved on the Internet. At the same time we have to face a large number of instant messaging contact resources which cannot be accurately matched.This thesis worked on a detailed study of instant message and information retrieval, and analyzed the Internet's potential demand and business model. Six space technology is applied as the general theory of the system. Q&A system and natural language processing theory is also applied as a logical adhesive. Distributed system and grid computing is made as the physical implementation approach. Our system is improved by data mining. Based on the above technologies, we build the intelligent IM search robot system.Our system has the functionalities, including smart dialogue, corpus's machine learning, personalized searching and user behavior analysis. It provides fast accurate resources positioning and instant information release services. The system has already been on-lined and achieved good results.