Text Case Retrieval Algorithm Based On Particle Size And Application Research

The similar problems have similar solutions, so people usually solve new problems based on the past experiences. However, more and more human experiences are stored as textual document in computers, handheld computers and other electronic products. So textual case based reasoning has become the people’s needs.However, existing textual case retrieval systems mostly are not common ones, and the retrieval efficiency and the speed are low. The same time, it is easy to ignore the orders between words, make the retrieval just be physical match, and be lack of the intelligence like people solving problems that the general system works from the perspective of word.It is the strong performance of human solving problems, that human can observe the world form a different granularity, and can be easily convert form one level to another. Therefore, one of an important research direction of the field of artificial intelligence is making the computers have the capacity which similar to human solving problems. The main purpose that introduces the granularity thinking of human beings in the textual case retrieval is to reduce the computational complexity, get better retrieval efficiency, and improve the retrieval speed. The main contents are:1) Studying about the representation of granularity of textual case. Describing the textual case with words and sentences based on the inherent characteristics of textual case, and the core idea of granular computing.2) Studying on the model of case retrieval based on multi-granularity. The paper describes the problem of textual case based retrieval with quotient space theory, and divides the processes of textual case base retrieval in to dividing the quotient space for user text case and searching. It can reduce the complexity of the retrieval that searching after dividing the quotient space for user textual case.3) Researching about dividing the granularity of the case base. Dividing the case base or cases into different granularity whit different size, to improve the speed and efficiency of dividing the quotient space for user textual case and the retrieval accuracy, and reduce the retrieval time-consuming. 4) Researching the retrieval algorithms based on multi-granularity. Study on the algorithm of textual case retrieval that going on from words and sentences two particle size based on existing researches. And5) Designing and implementing the system of textual case retrieval based on multi-granularity. According to the model of multi-granularity retrieval and the retrieval algorithms, the paper shows the functions of the system and detailed design, and achieves the system.Through the research, experiments prove the feasibility and effectiveness of the model of textual case retrieval based on granular computing and the algorithms. And the paper solves some problems that existed in textual case based retrieval.