Research On Information Retrieval Based On Rough Fuzzy Set

With the rapid development of Internet and the upsurge of the quantity of information resources,the crisis of Information explosion comes up. The massive information beyond the people's imagination far and away.and it doesn't appreciate with it's tremendous value.The useful information mingle with the unuseful information.How to disentomb the valuable information from the massive information is a difficult problem that people must be faced with in the current world.It's also one of the hottest problems that people pay attention to in the field of intelligent information processing.Information retrieval is one of the hottest problems that people pay attention to in the field of intelligent information processing.lt takes up with organizing and storing information followed a certain pattern.and finding out the specific information according to the need. Fuzzy set theory and Rough set theory are continuation and inportant development of the classical mathematic set theory. They all can deal with incomplete and unversed information. information processing technology based on fuzzy and rough set theory is applied widely right now.Combining with these theorys can heighten their ability in the field of intelligent information processing.The information retrieval process includes three parts:How to express the documents,How to express the queries,How to make the documents and queries match for the purpose of retrieval and rank the result.In my research work,a new information retrieval model is designed using fuzzy set and rough set.It makes use of the inclusion relation between fuzzy sets in fuzzy set theory——theorem of include degree, we use the inclusion relation between fuzzy sets to make the documents and queries match,also we use the include degree to rank the documents that retrieved.It also makes use of the equivalence relation between keywords.this can reflect the relation between keywords,so it can search the Synonyms.Compared with traditional information retrieval model,this model resolves the shortcoming of strict match in Boolean Model and neglecting the relation between keywords in Vector Model. Compared with traditional Rough set Model,this model can express the documents using weighting that means how important these keywords are. Also through the queries,you can express your interest using weighting of keywords.my research work includes details as following:(1)Analyze several kinds of traditional information retrieval models,such as Boolean Model, Vector Model, Probabilistic Model.Research their merit and demerit.(2)Introduce related knowledge of Rough set,Fuzzy set. We use theorem of include degree to information retrieval,then form a new information retrieval model based on Fuzzy set.This model resolves the shortcoming of strict match in Boolean Model and only using Boolean quantity expresses keywords.(3)Consider the shortcoming of neglecting the relation between keywords in the information retrieval model based on Fuzzy set,we come up with the new model.This model is based on Rough Fuzzy set.This research uses the theory of Upper Approximation, equivalence relation of synonymy.Using the knowledge.It can extend the synonym in the queries and docements.That means we can use the theory of Upper Approximation, equivalence relation of synonymy to extend the information of the queries and docements.It can increase the recall ratio.(4)Based on above-mentioned idea,we conduct the experiment to do matching simulation,then compare the model based on Fuzzy set and the model based on Rough Fuzzy set.From this,it can reflect the advantage of the model based on Rough Fuzzy set.However, the experiment of this thesis is only based on the matching simulation,so more researches will be done,For example,How can applied it in practical application.This is also the most important topics for future research.