With an increase in amount of data also the speed of access to information has increased. Large data is the distorted
form of every data well gained from different sources such as photos, , videos, social media sharing, network blogs, log files, etc.
into a consequential and feasible forms. Clustering methods are very useful. Clustering process permits very analogous data to
be positioned under a group by unscrambling the data into a explicit group. Once datasets are separated, outlier detection is used
to discover false data. In this research work, goal is to formulate data clustering and outlier detection process quicker by using
MapReduce technology with modified K-means clustering method. Clustering on enormous data can be time overriding. Hence,
MapReduce computing design is utilized and focused on reliable, unfailing and swift clustering process by this technology. The
successful execution with comparison to traditional K means is drawn. The results are offered in tables and graphs using sample
dataset.
