Font Size: a A A

Expert-Application Grouping And Matching Algorithm And Projects Application Management System

Posted on:2010-05-04Degree:MasterType:Thesis
Country:ChinaCandidate:H LiuFull Text:PDF
GTID:2178360278952294Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
Objects grouping and matching simultaneously is an important means in grouping heterogeneous objects. Project applications management is an ordinary work in today's society. During project-review, grouping and matching experts and applications is a typical grouping and matching problem.Objects grouping and matching is similar to objects co-clustering. In the co-clustering using Bipartite Spectral Graph Partitioning, the relationships between heterogeneous objects are modeled as a bipartite graph at first, and then the bipartite graph is partitioned under partitioning principles. The partitioning problem of the bipartite graph can be posed as a matrix decomposition problem according to the spectral graph theory. At last, the heterogeneous objects are clustered through mapping to the clusters of the eigenvectors witch are clustered by k-means algorithm.In order to group and match the experts and applications in project-reviewing, the Expert-Application Algorithm is proposed in this paper, which is based on the co-clustering using Bipartite Spectral Graph Partitioning. In this algorithm, the number of the expert and application are configured at first. And then, the relationship between experts and applications are depicted as a bipartite graph model. The edge-weight is calculated by the relationship weight calculating formula witch is proposed in this paper. After decomposing the matrix witch is formed by the edge-weight, we get eigenvectors. At last, experts and applications are grouped and matched through mapping to the groups of the eigenvectors which are created by k-means grouping algorithm.The results of experiments show that Expert-Application Algorithm can group and match experts and applications. And the grouping result is correct and reasonable.In order to improve the efficiency of projects applying, a projects application management system is designed and developed in this paper. This system consists of four subsystems: project application, management of department in charge, management of reviewing department, project reviewing, and user logging control.
Keywords/Search Tags:Grouping and Matching, Bipartite Graph, Graph Partitioning, Projects Application Management
PDF Full Text Request
Related items