Location Prediction for Social Media Users Based on Information Fusion

The real locations of social media users have always been a hot spot for people. However, considering personal privacy and other factors, most locations provided by users are ambiguous, missing or wrong. In order to get users’ real location, we collect various types of geographic related information from users in social networks and propose an information fusion network model to organize the information efficiently. After that, we take advantage of the iterative-based information fusion method to process the geographical related information in the information fusion network and the outputs are used as users’ geographical location. Finally, the experimental results show that our research method can greatly improve the prediction accuracy and reduce the corresponding distance error.