Secured Authorized Recommendation (SAR) Analyzing the User Relationships on Social Media through Machine Learning (ML) Algorithm

Abstract

With the growing social media more challenges and security have been a big concern in today’s’ world. Users usually have multiple identities (IDs) in different social media websites with some other names and accounts. Particularly, kids and teenagers are affected due to several unwanted recommendations of sites, games or videos. The primary patterns of the online user-relationship are being analysed in the proposed system. The proposed system also tries to discover the process through which the user relationship network functions. Also, the user perspective can be tracked and recommendations are made according to the users requirements. Moreover, the system implies a secured Authorized Recommendation (SAR) theory to confirm the age limit of the users from the parents/guardians. The system suggests recommendation according to their age seeking permission of the parents. To implement the theory the system accommodates a   social user relationship management scheme. Through machine learning algorithm (ML) and Keyword matching the user relationships are studied validating the personalized accounts of the users.  For Experiment analysis a twitter dataset from kaggle is taken and the user relationship analysis is made and suitable recommendations are made accordingly. The experiments prove the proposed system to be more efficient.