Open AccessUnderstanding the Users Personal Attributes Similarity Across Online Social Networks
Author(s) -
Waseem Ahmad,
Asif Rashid
Publication year - 2019
Publication title -
international journal of innovative technology and exploring engineering
Language(s) - English
Resource type - Journals
ISSN - 2278-3075
DOI - 10.35940/ijitee.k1297.0981119
Subject(s) - computer science , world wide web , the internet , similarity (geometry) , security token , social media , internet privacy , matching (statistics) , personally identifiable information , social network (sociolinguistics) , information retrieval , artificial intelligence , computer security , mathematics , statistics , image (mathematics)
In this modern era of technology, everyone accessing the Internet is obsessed with social media. A User accesses different social media services to fulfill his diverse needs. For instance, Instagram is mainly used for sharing personal visual content while Twitter is known for finding latest news and trends, similarly Facebook for personal posts. Such services lead to the distribution of personal information of an Internet user on these platforms. In this paper, we build a framework to discover the relationship among the attributes of a user across the social media.We use different fuzzy string matching algorithms to find the similarities between the attributes. We extract the ‘name’ and ‘username’ from a publicly shared dataset and apply two character based and token based algorithms on these features. The results are indicative of the fact that only a limited number of users share the same name and username across the sites. On further analysis, it is found that although name and username of most of the users do not exactly match, they tend to be similar with the infinitesimal difference like; underscore, period, one digit numbers, etc. This study provides an analysis of the typical variations in names and usernames, which can further be studied for the extension to other social networks This profile will help in behavior analysis of a user, which will further help us to improve recommendations and analyze for criminal behavior and similar applications.