The Study Of The Physical Chain Of Chinese Film Reviews Based On Knowledge Maps

With the rapid development of the Internet,the amount of data is gradually becoming larger.It is critical to extract useful information from a wide range of Internet data and make it a human intelligence helper.Extracting information from natural texts is a matter of natural language processing,and entity linking is one of the important technologies.It has the potential for artificial intelligence Q & A system,information extraction and retrieval,machine translation and so on.The purpose of the entity linking refers to the objective matter identified by the name of the entity being investigated,which is divided into two parts: the candidate entity generation and the candidate entity.There is little concern about candidate entities,and most of them focus on how to achieve candidate sorting.Aiming at the problems existing in the existing research,this paper presents a entity linking algorithm based on knowledge graph in Chinese film review.The main contents are as follows:1.This paper presents a knowledge graph model based on ontology.The basic elements of knowledge graph namely concept,relation,rule are defined respectively,and the semantic relation between concepts is embodied by the concept,relationship and rule,and then forming knowledge graph among concepts.By using the knowledge graph to carry out the entity chain,the steps of candidate entity generation are omitted.2.The algorithm improves the traditional concept similarity model,and fully considers the length of the connection path of each concept node in the knowledge graph model,the depth of the node and the density of the node.The relationship between the concept nodes is weighted by the conditional probability.The improved model is used to calculate the semantic similarity between the candidate entities and concept nodes,and entity which has the max value of similarity is the target entity.Combined with the CCKS-2016 national knowledge graph and semantic conference evaluation task,this paper processes the given knowledge base firstly,and then describes the realization process of the entity linking algorithm based on the knowledge graph.The results show that the accuracy rate of the entity linking is as high as 89.2%,it is proved that the algorithm of knowledge graph modeling and the entity linking based on knowledge is effective and has high accuracy.