Vertex Unique Labelled Subgraph Mining for Vertex Label Classification
A mechanism is presented to classify (predict) the values associated with vertices in a given unlabelled graph or network. The proposed mechanism is founded on the concept of Vertex Unique Labelled Subgraphs (VULS). Two algorithms are presented. The first, the minimal Right-most Extension VULS Mining (minREVULSM) algorithm, is used to identify all minimal VULS in a given graph or nework. The second, the Match-Voting algorithm, is used to achieve the desired VULS based classification (prediction). The reported experimental evaluation demonstrates that by using the minimal VULS concept good results can be obtained in the context of a sheet metal forming application used for evaluation purposes.
KeywordsData mining Graph mining Vertex unique labelled subgraph mining Classification
Unable to display preview. Download preview PDF.
- 4.Liu, W., Liang, Z., Huang, T., Chen, Y., Lian, J.: Process optimal ccontrol of sheet metal forming springback based on evolutionary strategy. In: 7th World Congress on Intelligent Control and Automation, WCICA 2008, pp. 7940–7945 (June 2008)Google Scholar
- 6.Salhi, S., Coenen, F., Dixon, C., Khan, M.: Identification of correlations between 3d surfaces using data mining techniques: Predicting springback in sheet metal forming. In: Proceedings Proc. AI 2012, pp. 391–404. Springer, Cambridge (2012)Google Scholar
- 7.Yan, X., Han, J.: gSpan: Graph-based substructure pattern mining. In: Proceedings of the 2002 International Conference on Data Mining, p. 721 (2002)Google Scholar