Understanding the Users Personal Attributes Selection Tendency across Social Networks

Today, we are using social media services for virtual interaction with friends and colleagues. With each passing decade, several new and interactive social networks are coming into the existence to serve the people. Each established social network emulates the trait of new media and add the new features to survive in the versatile market. To enjoy the services of newly introduced social media, people migrate from one social network to the other by sharing her personal identities in a variety of formats. To understand the users' migration pattern from one social media to another, here we propose a hybrid personal information retrieval model consisting of two sub- methods, namely; cross-link based personal attributes matching and cross-platform based common friend relationship. Both the methods are effective for finding the similarities among the profile attributes across Twitter and LinkedIn. Further, both the techniques uses Levenshtein distance to match the user's personal attributes across the social media. The result shows that accuracy of name similarity using the first method is 66%, while the username and name similarity is 20%. Second method shows that 11.2% users contain the same or similar username and name on both the sites. It is found that the accuracy of the cross linked based approach is better than the friend relationship based approach.