A Mapreduce Computing Base Fuzzy Classifier Extraction for No-SQL Data Classification

摘要

In current literature of data science one can found that the immense of efforts are paying for scaling conventional data mining algorithms on MapReduce framework to make them capable of handling massively gathered data commonly being referred as Bigdata. But the one important factor most of those works are missing is that the majority of data in Bigdata bases being stored in no-SQL format to accommodate variety of data gathered from heterogeneous sources. On the other hand associative classification techniques and their fuzzy logic extensions which are popular due to their classification accuracy and interpretability have been successfully scaled up on MapReduce framework to handle huge volume and uncertainty of big data. Being focused up on importance of data variety and understanding impact of parallel fuzzy associative classification technique in efficient handling of volume, veracity and variety properties of Bigdata bases this work proposes a MapReduce computing solution for extracting fuzzy associative classifier from data stored no-SQL formats.