Researched Information Retrieval Based On Bayesian Network

On the base of statistics, the bayes network is a method of data mining. In essence the bayes network is a directed acyclic graph presenting directly the reliance relations among many variables. It depicts the cause and effect relations by a directed acyclic graph and the chummy relations by a conditional probability distribution table among all nodes. Moreover, we can incorporate the prior knowledge into current data effectively and get a more reasonable result. Especially when the current data are scarce or hard to obtain, the advantage of the bayes network is evident.With the rapid development of Internet technology and then the surpassing increase of all kinds of information characterized by geometric progression, the intellectualized information retrieval has become a major research topic, for the traditional information retrieval can't meet the requirements of users. The function affecting a retrieval systematically has many factors, most key's be still the information retrieval model. The information retrieval model efficiency has decided entire information retrieval effect.In this paper, we introduced three classes of mathematics models of information retrieval, which are set model, algebra model and probabilistic model. The characteristics of these models were also analyzed. And have analysed several advantages of the information retrieval making use of bayes to be in progress coming the network: bayes network method has the solid rationale; bayes network has the mature probability speculate algorithm and exploitation software; bayes network is more suitable to in information retrieval model; bayes network has the very strong learning ability. And union information retrieval self characteristic , we designed a bayes network model on inference network model basis. We made some improvements in bayes network of information retrieval , setted a limit for by the fact that probability in model is in progress to bayes network, have facilitated calculative amount of work. At the same time, when consumer is importing the inquiring keywords, have variety of cause sometimes because of oneself, but not enough accurate , not enough meticulous, to have affecting an information retrieval result gravely. For the problem resolving this, we have expanded in making use of mining association rules method to have been in progress to searching a word once again based on the bayes network. Such can resolve the inaccurate consumer input inquiry keyword problem effectively.Finally, by the fact that the experiment draws on recall rate and precision rate , we comparing the information retrieval model brought forward by us, and the other three kinds tradition information retrieval model . That result testifies information retrieval model brought forward by us is very effective.