How to implement a big data clustering algorithm: a brief report on lesson learned

The current era ofBig Data has forced both researchers andindustries to rethink the computational solutions for analyzingmassive data. In fact, a great deal of attention has been devoted tothe design of new algorithms for analyzing information availablefrom Twitter, Google, Facebook, and Wikipedia, just to cite a few ofthe main big data producers. Although this massive volume of datacan be quite useful for people and companies, it makes analyticaland retrieval operations really time consuming due to their highcomputational cost. A possible solution relies upon the possibilityto cluster big data in a compact but still informative version ofthe entire data set. Obviously, such clustering techniques shouldproduce clusters (or summaries) having high accuracy. Clusteringalgorithms could be beneficial in several application scenarios suchas cybersecurity, user profiling and recommendation systems, tocite a few.