Twitter Analysis for Intelligent Transportation

Abstract

Abstract The amount of data available online has grown enormously over the last decade as a result of the rapid growth of smartphone users and the availability of communication applications. Due to the anonymity and instantaneous nature of social media broadcasting compared to conventional attitudinal survey methods, social media mining is becoming popular for complementing traditional traffic detection methods due to its accessibility in reaching a large population and the opportunities for reflecting the true and immediate behaviour of participants for free. This study presents a framework for Arabic Twitter content analysis to gain transportation insight. The study is done with a dataset of more than 1 million tweets collected within 3 months. The proposed model comprises three main components: data acquisition, data analysis and the reverse geotagging scheme (RGS). The RGS tackles the problem of lack of location information in the tweets. Results show that 13% of the dataset reports traffic-related incidents with an overall precision of 55% and 87% for incidents identification prediction without and with reverse geotagging, respectively. This proves the efficiency of the developed analyser in identifying tweets on transportation and the potential of the RGS in defining the location of tweets with no registered location information.