Real Time Event Location Detection Based Mobility Pattern Modelling for Social Media User Mobility Analysis

-Detection of important events in a region through social media has been a recent development with scope for multiple applications. One of the applications is the analysis of the mobility of the users in the region to adapt the energy and resources as per their movement. For this purpose, the travel or movement behaviour of the social media users is analysed through their posts and related messages. An efficient method of mobility pattern modelling named as Event Location based Mobility Pattern Modelling WebClickviz (ELMPM-WebClickviz) is proposed in this paper based on the event location detection. Initially the social media data are collected and preprocessed. Then the geographical as well as temporal information are extracted along with the time and distance parameters. Then the pattern modelling is initiated using Sequential Hierarchical pattern clustering which detects the continuous events from the user data along with the location of occurrence. Based on these results, the mobility behaviour can be modelled with higher accuracy. The evaluation results prove that the proposed model provides efficient mobility patter modelling to be utilized for the organizations and official concerns in fulfilling the resources and needs of the users of that location. Key-Words: Social media, Travel behaviour, Event location detection, Mobility pattern modelling, WebClickviz, Sequential Hierarchical pattern clustering