Volume 1, Issue 3

[1]-Big Data Analytics: Map Reduce Function using
BIRCH Clustering Algorithm

Author Details: S.Swarnalatha , K. Vidya
CSE,JNTUH, Hyderabad, India
It is well known that, in Big Data information is represented in unstructured form and NoSQL is used for query processing. The volume of data also too large and simple Query processing is not sufficient and irrelevant. From that large volume of data, extracting the knowledgeable information is a big challenge. To analyze that, various Big Data analytical techniques are available in the market, that uncovers hidden patterns, market trends, customer preferences and other useful information that can help the organization to take useful decisions within less amount of time. For such applications, Map reduces frame work has recently attracted, that was introduced by APACHE HADOOP. However, conventional Map-Reduce function uses K-Means Clustering algorithm, which will work efficiently on numerical data only with high time complexity. This paper gives an idea how BIRCH works efficiently on large databases concerning running time, space required, quality, the number of I/O operations applied. It shows linear scalability with respect to a number of objects.

[Download Full PDF] [Page 01-07] DOI: 10.6084/m9.figshare.5350141