The Research And Implementation Of Building Semantic Web Based On Probability And Statistics

We can see the resources stored in the network in the World Wide Web of real-life. A computer is just a bridge to connect the various web pages, sometimes the computer is not complete in accordance with the needs of users to search for information, and in many cases the contents of the search is not users which want. The emergence of semantic web is just solving this problem. It enables the resources of the computer searching is just the thing which users want to find, and provides the users a more accurate searching result. So the semantic web is concerned by more and more people. Therefore, relevant research of building the semantic web has become a hot spot.Hidden markov model (HMM) has been more and more applicated in the computer. Human often summs up the laws of its development from the life. HMM uses the available datas to deduce the future, and learns the available datas to suggest the existence of the conclusions established.In this paper, it proposes a method to build semantic web based on probability and statistics methods through the research of these two technologies, this method combines them. The main application part of probability and statistics is hidden markov model, the three algorithms of hidden markov model is applied in the corpus, the computer to automatically build the semantic web. The use of hidden markov model calculates the state transition probability matrix between these words, and it builds directed acyclic graph with the probability to use the probability value calculated, the various values has been trained for building the semantic web model. At the same time of using owl representation, it also fully takes into account the hierarchical relationships between the concepts. It has been built semantic web using a directed acyclic graph automaticly.