Classifying Vehicle Traffic Messages from Twitter to Organize Traffic Services

Abstract

Communication in social networks as Twitter, Facebook, and Line application is a channel telling a traffic status by a text message. Traffic services organization by peoples monitors messages from the social network in order to take actions when having the accident. This research focuses on creating a system that classify the text-message from social networks, to get the status of traffic, type of accident, level of accident and GPS location in order to consider and take action by the traffic services organization. The designed system in this phase uses text analysis by considering structure and message styles and the fully design applies hidden Markov models to classify the type of the message. Short message from the social network came from Twitter, JS100-radio and FM91.