Social Media: A Ingenious Tool To Assist Users

1101 Words3 Pages

The research paper Zafrani et al cite{Zafarani:2013:CUA:2487575.2487648} introduces a method known as Modelling Behavior for Identifying Users across Sites (MOBIUS). The information shared by the users on Social media site provides a social fingerprint that helps in the identification of the users. MOBIUS studies this information that is provided by the users like username, gender, habits, positions, roles and sex on different Social media sites to identify certain unique behavioral patterns resulting in some repetitive information. This redundancy information helps in learning the identification function. MOBIUS captures this repetitive information in terms of data features. We can learn the identification function by the tradition of data mining research and machine learning. To achieve this the authors propose a supervised learning framework. This framework uses prior information of an individual and the data feature set captured in MOBIUS. The supervised learning framework implements the MOBUIUS method and provides us with a unique behavioral pattern of the user along with the individual function associated with it. MOBIUS is performed via the data mining techniques and they are either regression or classification, and it uses knowledge discovery to generate a probability function. It assists in selecting the most likely individual who owns the username. MOBIUS contains the features constructed to capture information redundancies which arises due to these patterns and a learning framework to generate a unique behavioral pattern of the user. This uniqueness helps in the identification of users across social media sites. The value of the individual function we get can also be used to solve the problem of age verification, protecting the youth using internet from the bushwacker, to prevent crime, identify criminals, etc.

The research paper Bart et al cite{Stajner:2013:ASS:2487575.2487659} proposed a method that uses Greedy Iterative Sampling algorithm to select a subset of messages from a given sample of messages posted by users across Social media sites. This algorithm takes a collection of messages refereing to a given news article, the sample size of the message and returns a sample set of messages. The algorithm is based on the idea of mining the sample messages provided and then extracting a subset of messages enriched with the four message-level indicators and a set-level indicator. These indicators in the message represent the interestingness of the message for a given news article. The method then uses knowledge discovery techniques to select the most interesting messages to a reader in the situation of the news article.

Open Document