MergeGraphs: a web-based system for merging heterogeneous big graphs

The quantity and quality of the information represented by means of graph structures is more and more increasing nowadays. Merging this information assumes a deep relevance in order to identify unknown relationships and properties and conduct different kinds of analysis. However, this task is complicated due to the variety of formats with which graphs are represented and their size. In this paper we propose mergeGraphs, a web based system in which the user can graphically drag and drop heterogeneous graphs and specify transformation and merging operations (union, intersection and join). The obtained execution plan is translated in Pig Latin scripts and executed in a cluster of machines. The use of Google App/Compute Engines allow to develop a scalable distributed approach.