IDENTIFYING THE REAL USER IN THE SOCIAL NETWORK USING CLASSIFIERS

Neha Doggalli, R Senthil Kumar

Abstract: Social identity linkage is essential to connect different social media platforms. It is critically needed for business intelligence to investigate about the user profiler of the social media. To address this issue HYDRA framework is used. It has three steps mainly heterogeneous behavior, structure consistency and normalized margin based linkage. In heterogeneous behavior model, thebehaviorof certain attributes are matching against the information missing and the behavior similarities are described to know the person is real or not. In structure consistency model, the purpose is to construct the structure and behaviorregularity of users across different social networks thus the identity link can be performed on set of users. In normalized margin based linkage, it propose a function and learn the linkage function by considering multiple optimization where both overseen pair wise linkage function and structure consistency are conducted to get optimal solution. This model can find the missing information and helps to find out the linkage between the two platforms. This work summarizes about the identification of user profiles or the information using HYDRA

Keywords: Data Mining, Structure Consistency, Normalized Margin.

DOI: https://doi.org/10.15623/ijret.2016.0516094