A hierarchical multi-relational clustering algorithm based on IDEF1x

作者：Huang, Shao Bin; Cheng, Yuan; Wan, Qing Sheng; Liu, Guo Feng; Shen, Lin Shan

来源：Acta Automatica Sinica, 2014, 40(8): 1740-1753.

DOI：10.3724/SP.J.1004.2014.01740

• Hierarchical model
• Multi-relational clustering
• Multiple clusterings
• Re-utilization
• Real-world
• Shortest path
• Transfer results

There still exist some problems in multi-relational clustering, such as the primitive characters of data may be ignored when getting information reflected by one-to-many relationship through statistical methods, and information reutilization may result from the loops in the relationship between tables. We feel that a model can be rebuilt to solve these problems based on the characters of relationships in the IDEF1x model. Therefore, a hierarchical model of relationships between tables in multi-relational dataset is built based on the IDEF1x model, the effect brought by different relationships in the model as well as the integration of multiple clustering results are defined, and then a new multi-relational clustering algorithm is proposed based on these definitions. Finally, the experiments on real-world and synthetic dataset indicate the accuracy of our algorithm.

• 出版日期2014
• 单位哈尔滨工程大学