Abstract
SET COVER-RRK FEATURE SELECTION TECHNIQUE FOR ENHANCING THE ACCURACY IN SOCIAL NETWORK DATA
R.Rajkumar* and Dr. Anbuselvi
ABSTRACT
Online Social Network like Face book, Twitter, LinkedIn etc., have become the popular interaction, recreation and socialization facility on the internet. Users choose greater engaging sites, every time they will notice familiar faces like friends, relatives or colleagues. A “feature” or “attribute” or “variable” refers to an issue of data. Usually earlier than gathering data, features are detailed or chosen. Features may be discrete, continuous, or nominal. Generally, features are characterized as: 1. Relevant: There are features which have a power on the output and their position cannot be assumed by using the relaxation. 2. Irrelevant: Irrelevant features are defined as those features not having any influence on the output, and whose values are generated at random for every instance. Feature subset selection in Online Social Network can be analyzed as the exercise of identifying and removing of as lot of irrelevant and unnecessary features as achievable. This is for the cause that, irrelevant features do no longer make a contribution to the predictive accuracy. First shifting out irrelevant features from the Online Social Network data set[5], for irrelevant features are removed by using the features having the value above the predefined threshold. The reason of this research paper is twofold; Identifying and removing the irrelevant features in Online Social Network with latest solutions for Consistency Measure in an Online Social Network.
[Full Text Article]