Issue No.02 - March/April (2000 vol.15)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/5254.850825
Using databases represented as graphs, the Subdue system performs two key data-mining techniques: unsupervised pattern discovery and supervised concept learning from examples. Applications to large structural databases demonstrate Subdue's scalability and effectiveness.
Diane J. Cook, Lawrence B. Holder, "Graph-Based Data Mining", IEEE Intelligent Systems, vol.15, no. 2, pp. 32-41, March/April 2000, doi:10.1109/5254.850825