Online social Network Trend Discovery Using Frequent Subgraph (FSgM) Mining

Abstract

A social network (SN) is an online platform where people use to contact dynamically with various social networks and establish social relationship with other people and share similar personal, career interests, activities, and real-life connections.An OSN is a dedicated website or application used to communicate with each other. In today’s internet research world due to the advancements in computer hardware and efficient computing processing power, construction of large-scale graph data mining has become a plus point to graph mining researchers. So Graph mining has become one of the most important disciplines in the research area of data mining.A social network contains a huge amount of unstructured data resulting to complex challenging problem to many researchers. A Frequent Sub graph mining (FSgM) techniques is been used to identify the frequent pattern trends existing in the network for networks analysis. A Frequent Sub graph mining (FSgM). is playing a very prominent role in the core graph operation with its domain areas like social networking analysis, web data mining bioinformatics, graph data management, knowledge exploration and security. Frequent subgraph mining have extremely high computational complexity. Finding subgraph patterns for frequently reappearing social network graph dataset is important to identifying the interpretable structural properties of complex networks to study the trend and also for social balance and status findings. Many researchers have proposed many graph mining algorithms, but we observed a few are working for capturing the important element of Social networks, so this became trends to discovery of frequent pattern mining. In this paper we introduces a novel FSgM approach, called  gSpan (Graph Based Substructure Pattern Mining)    to identify the frequently occurring  pattern trends in  the social network data. We considered Facebook social media data for identifying the frequent pattern trend and performed analysis.