The Research Of Intelligent Search Engine Technology Based On Granular Computing

With the development of Internet, search engine already becomes one of the important means to obtain information in the people's study and work. Therefore, it's imperative to improve the efficiency of search engine, on the one hand, we need to popular the correctly way and skills on the use of search engines, on the other hand, to improve search engine's operate mode is also important.Artificial Intelligence Search is an important area of research, As we all know, the problems about artificial intelligence is usually uncertain, ambiguous, incomplete, massive and often can not find a conventional solving path, but if the scale of the problem is not too large, at this time we can use the exhaustive search strategy, such as breadth-first search and depth-first search, Because their relative calculation time consumption is easy to control, so it is widely used in small-scale expert systems.In the actual AI systematic application, the solution scale of problem finding very big and available scheme is also more. Therefore, it can not to obtain our aim if we still use a simple exhaustive search strategy, at this time, heuristic search by using empirical knowledge to guide the direction of the next search may do our favor. This way is not only reducing the Possibility of blindness seaching, but also improving the efficiency. Although heuristic search is advanced, it still has not overcome the explosion problem in the computing index.Granular Computing is theory with the target at solving the complex problems of artificial intelligence, it belongs to a branch of soft computing science, it's a new concept and paradigm, covering all of the theory about methods, techniques and tools of research, its ideological essence is simple demand, to meet the approximate accuracy of exact solutions with low-cost alternative, using imprecise, incomplete, uncertain, and vast amounts of information to achieve the intelligent system control easily.This paper attempts combine the theory of quotient space with intelligent search engine, raised the Granular Particle Clustering and the Granular Statistical Heuristic Search Algorithm, make discussion of theory and research focusing on the granularity of the data. These two algorithms are all make full use of hierarchical to divide particle size. Based on the hierarchical, it reducing amount of unnecessary operations and the computational complexity. At the end of this paper, These two algorithms was verified by a experiment, they applied to a category of its scattered paper library on the reconstruction of the data repository and retrieval,based on this theory,we can obtained a better search algorithm than the classical results.