| With the rapid development of economic and social information,the types of information resources are increasing rapidly,and the competition between enterprises can not be separated from these information resources.Enterprises Through information technology to build a variety of information transfer and sharing platform to promote the development of enterprise information,and gradually improve their market competitive position.Because of the dispersion of information system development,resulting in enterprise internal information can not be real-time sharing,formed"information isolated island";secondary development or the introduction of new system resulting in data integration very difficult,to well solve these two problems must be reintegrated planning of enterprise information,information resource planning.Thematic database planning is an important part of information resources planning.This paper mainly studies the subject database planning from the following aspects.The main application of business analysis stage Maximum Spanning Tree Clustering Fuzzy Mathematics to complete the separation of functions of the domain.Relying on the experience of business people select the best division threshold value,based on the application of statistics in the F statistic calculation formula assistance.After experimental verification,F statistic can well calculate the optimal threshold,the optimal number of clusters,cluster and then get the best results,then do a small functional domain into the results obtained can be adjusted according to the actual needs of the ideal.In the F statistic assisted to improve the functions of the domain into the speed and quality.Gathered in the entity analysis phase,through investigation of the actual situation and found that the entity-active matrix can not accurately measure the relationship between entities and activities proposed entities-activity frequency matrix,followed by proposed entity-percentage of activity matrix.During physical similarity calculation,cosine similarity formula is applied,followed by the proposed generalized concept of similarity and quality assessment relating to database planning function.Finally,this paper entity clustering algorithm.At last,taking the University as the actual application background,the theme data-base planning is completed.Based on the new method presented in this paper,the subject database planning of the university has been well planned,and the numerical comparison of the quality evaluation function has verified this conclusion. |
