In contrast to mining over transactional data, graph mining is done over structured data represented in the form of a graph. Data having structural relationships lends itself to graph mining. Subdue is one of the early main memory graph mining algorithms that detects the best substructure that compresses a graph using the minimum description length principle. Database approach to graph mining presented in this paper overcomes the problems – performance and scalability – inherent to main memory algorithms. The focus of this paper is the development of graph mining algorithms (specifically Subdue) using SQL and stored procedures in a Relational database environment. We have not only shown how the Subdue class of algorithms can be translated to SQL-based algorithms, but also demonstrated that scalability can be achieved without sacrificing performance.
CITATION STYLE
Chakravarthy, S., Beera, R., & Balachandran, R. (2004). DB-subdue: Database approach to graph mining. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 3056, pp. 341–350). Springer Verlag. https://doi.org/10.1007/978-3-540-24775-3_42
Mendeley helps you to discover research relevant for your work.