Parallel K-prototypes for Clustering Big Data

摘要

Big data clustering has become an important challenge in data mining. Indeed, Big data are often characterized by a huge volume and a variety of attributes namely, numerical and categorical. To deal with these challenges, we propose the parallel k-prototypes method which is based on the Map-Reduce model. This method is able to perform efficient groupings on large-scale and mixed type of data. Experiments realized on huge data sets show the performance of the proposed method in clustering large-scale of mixed data.