Finding Frequent Subgraphs and Subpaths through Static and Dynamic Window Filtering Techniques

Big data era has large volumes of data generated at high velocity from different data sources. Finding frequent subgraphs from the graph streams can be a challenging task as streams are non-uniformly distributed and continuously processed. Its applications include finding strongly interacting groups in social networks and sensor networks. To find frequent subgraphs, we proposed static single-window technique and dynamic sliding window techniques. We also proposed enhancements by extending proposed static approach with its variations and extending dynamic approach in variations of incremental strategy to find frequent subgraphs. We also solved the sub problem to extract frequent subpaths from sequence of paths. Its applications include finding congested sections in traffic analysis. We applied our proposed static and dynamic techniques to extract the frequent subpaths from sequence of paths. We experimented the proposed dynamic and static approaches with real and benchmark datasets.