Clustering of social networking data using SparkR in big data

Abstract

Due to every day growing amount of data and changing the formats, the storing and management of these data is the challenging task for the organizations. Not long ago, datasets contained thousands of data items. Currently, different technologies can store, manage and process data with increasing volumes of unstructured and heterogeneous data, data of this type are known as Big Data. Big Data is the period for a group of such huge and complicated datasets that makes it problematic to store, manage and process with existing data processing tools. Now, in Big Data, maximum of the data created is not structured. Therefore, the new situations imposed by Big Data present grave challenges at multiple levels, together with clustering problem of these data. Clustering is one of the significant Big Data analysis problems, where very large amount of heterogeneous and unstructured data must be grouped together. Here we have describe the k-mean and hierarchical clustering methods; great attention to k-means method lends itself because it remains one of the most sought-after other approaches and it is also implemented in innovative technologies for analyzing Big Data. This paper describes different categories of data, the management of unstructured data in Big Data and the clustering analysis of social network data using SparkR.