Research On Mutual Enhancement Of Entity Resolution And Schema Matching In Web Information Intergration

As the Internet develops rapidly,WWW has become a large information and knowledge library. However, the information in it is too numerous and complicated to use. And then Web information integration comes into being, which can analyze, filter, integrate the Web information, and provide a unified knowledge overview and access interface(mode), and Web resources can be used efficiently. A Web information integration system can be divided into several different but closely related components:Domain Model Construction, Data Sources Process, Data Extraction, Schema Matching, Entity Resolution and User Application Interface construction. Although Schema Matching and Entity Resolution are the two important components of the Web information integration system, they are most commonly studied separately.How to efficiently realize the mutual promotion of Schema Matching and Entity Resolution which are originally separate parts is mainly studied in this paper, and some correlation algorithms are proposed, combining with the traditional Schema Matching and Entity Resolution technology. The main research includes:1.An idea that schema matching can accelerate entity resolution is proposed, and the correlation algorithm is also put forward.In the Web information integration process, most schema model we have obtained are with the instance model.If the instances in two different modes point to the same entity, the value of schema matching property in the two schema modes which is corresponding to the entity should also be similar. A specific and feasible algorithm is also put forward and validated by experiment.2. An idea that entity resolution can accelerate schema matching is proposed, and the correlation algorithm is also put forward.In the Web information integration process, when we get the need for a entity resolution, if we know the corresponding attributes of the schemas, then the two instances of a unified entity in the corresponding value of the property should also be similar. A specific and feasible algorithm is also put forward and validated by experiment.3.An idea about the mutual promotion of Schema Matching and Entity Resolution is proposed, and the correlation algorithm is also put forward.Base on the research ahead, an idea is proposed naturally, that is, realizing the mutual promotion of entity resolution and schema matching and improving the performance of both. We also give a specific algorithm, and the mutual promotion of entity resolution and schema matching is realized by an iterative way. Experiments show that the algorithm can improve the performance of entity resolution and schema matching efficiently.Schema matching accelerated by entity resolution mainly promotes a new idea in the area of pattern matching, while entity resolution accelerated by schema matching mainly promotes a related method In connection with the area of entity resolution. Schema matching and entity resolution are essential components of Web information integration system. In this case, the acceleration between both sides is of great significance. The purpose of research on mutual enhancement of entity resolution and schema matching is to improve one's performance by result of the other in the integration process, which is helpful to the entire Web information integration system.Exploratory study on how to effectively realize the mutual enhancement of entity resolution and schema matching in Web information integration is done in this paper, and we hope that an effective idea and method to this issue has been provided.Research in this paper is based on the technology which is applied widely in the current Web information integration field,which not only provides new ideas for promoting Schema Matching and Entity Resolution, but is helpful in promoting the Web information integration system. This makes the research in this paper appears not only exploratory theoretical research value, but practical significance and value.