Large Scale Graph Mining with MapReduce: Counting Triangles in Large Real Networks

In recent years, a considerable amount of research has focused on the study of graph structures arising from technological, biological and sociological systems. Graphs are the tool of choice in modeling such systems since they are typically described as sets of pairwise interactions. Important examples of such datasets are the Internet, the Web, social networks, and large-scale information networks which reach the planetary scale, e.g., Facebook and LinkedIn. The necessity to process large datasets, including graphs, has led to a major shift towards distributed computing and parallel applications, especially in the recent years. MapReduce was developed by Google, one of the largest users of multiple processor computing in the world, for facilitating the development of scalable and fault tolerant applications. MapReduce has become the de facto standard for processing large scale datasets both in industry and academia. In this Chapter, we present state of the art work on large scale graph mining using MapReduce. We survey research work on an important graph mining problem, counting the number of triangles in large-real world networks. We present the most important applications related to the count of triangles and two families of algorithms, a spectral and a combinatorial one, which solve the problem efficiently.