Using Web Log Files the Comparative Study of Big Data with Map Reduce Technique
Mudassir Khan, Shakila Basheer
Abstract
This research paper, discusses, how the implementations of a planned algorithm will be processed. The data set is collected from one of the engineering colleges in India. The Datasets are collected from different sources and how the raw data is a preprocessor. The data set is analyzed by using the Map Reducing Technique and that should carry out in naive byes and KNN algorithm. The Map-Reduce technique is parallel to computation using the key/value pair. Thus, the Map-Reduce system helps to analyze the information which will give the data of the potential clients, such as, login time, credit value, and a lot more at least reaction time. Finally, execution estimation has been done by using ROC bend to discover and affirm which algorithm is best similarly as orchestrating the weblog records per the general time, the specific time they have taken on a specific website. The Naive Bayes algorithm if its anticipated characterization. Like that the same process in the KNN algorithm. It is basic to perceive which the greatest classifier method is. To carry out this, the foreseen outcomes are approved during (ROC) Receiver Operating Characteristic curves.