Data Stream Partitioning Algorithms for Big Data Analytics: A Review

As technology advanced, it opened new ways of continuous data gathering. In several applications like network monitoring, Walmarts are creating huge volume of data, so it is not possible to store such high volume, multidimensional data on physical storage medium i.e. available to analyze only once. Various existing mining techniques will not be efficient for such type of streaming data. So these existing data mining techniques need to be enhanced for processing big data streams. This paper takes a critical review of each of the four types of stream clustering algorithms and concludes with some critical discussions, advantages and disadvantages of each type of algorithm as well as gives some future research directions.