Distributed Frequent Subgraph Mining Using Gaston and MapReduce

Abstract

This paper addresses this issue and devises a new method for frequent subgraph mining in order to retrieve the valuable information from the database that captured the attention of the users. This paper proposes the recurrent-Gaston (R-Gaston) algorithm for the frequent subgraph mining process by enhancing the existing Gaston algorithm. Moreover, the method uses support measures based on the frequency and page duration parameters in order to define the support for the proposed R-Gaston algorithm. The simulation of the proposed R-Gaston is carried out using the weblog and the MSNBC databases. The proposed R-Gaston has attained values of number of structures mined and the execution time as 184, and 1282ms for the MSNBC database, with 60 and 75ms for the weblog database, respectively.