The Research on Subject Database Division Based on Entity Similarity

摘要

The division of subject database, which is a major task of information resource planning, is the important way to realize the information sharing by solving the "Information Isolated Island" problem. The entity similarity, which makes the relations among entities more clearly, is presented by researching the existing methods of subject database division. Meanwhile, the subject database is divided in terms of the cluster analysis methods, which can conquer the bi-direction problem appeared in the process of aggregation analysis. Finally, an evaluation function is put forward to assess the results of division so that the division of subject database is more reasonable.