Paper abstractMining Frequent Connected Subgraphs Reducing the Number of CandidatesAndrés Gago Alonso - Advanced Technologies Application Center (CENATAV), CubaJosé E. Medina Pagola - Advanced Technologies Application Center (CENATAV), Cuba J. Ariel Carrasco-Ochoa - National Institute of Astrophysics, Optics and Electronics (INAOE), Mexico J. Fco. Martínez-Trinidad - National Institute of Astrophysics, Optics and Electronics (INAOE), Mexico Session: Graph Mining Springer Link: http://dx.doi.org/10.1007/978-3-540-87479-9_42 In this paper, a new algorithm for mining frequent connected subgraphs called gRed (graph Candidate Reduction Miner) is presented. This algorithm is based on the gSpan algorithm proposed by Yan and Jan. In this method, the mining process is optimized introducing new heuristics to reduce the number of candidates. The performance of gRed is compared against two of the most popular and efficient algorithms available in the literature (gSpan and Gaston). The experimentation on real world databases shows the performance of our proposal overcoming gSpan, and achieving better performance than Gaston for low minimal support when databases are large. |