USER LOCATION PREDICTION USING HYPERGRAPH IMPACT FACTOR IN TWITTER WITH GLOBAL DATA COMMUNICATION

Abstract

Twitter is one of the most prominent online media that acts as a global network for sharing sensitive real-time information like earthquake alerts, political news, product review, personality identification, criminal detection etc. along with regular usage, which is why knowing the location of a user in twitter gets at most important even though they do not tend to disclose it. In this paper, we propose a technique to detect the name of the locations for the twitter users. This technique involves a hypergraph-based map-reduce concept to represent the user tweets with their locations. The Helly property of the hypergraph was used to remove less potential words and the Impact Factor measure (IF) was introduced to calculate the score of each location for a particular user. The algorithm (HIF) was implemented in a big data environment provided by Hadoop and found to give an average accuracy of 78% which is well ahead of the existing methodologies. This method gives appreciable results, with high values of precision and recall for all locations.