Road Traffic Question Answering System Using Ontology

Abstract

Many people use social media to report and receive road traffic information, e.g., car accidents and congestions. We have implemented a Twitter-based traffic-related information reposting (retweeting) system, which users usually referred to as @traffy. To improve on our works, we propose an ontology-based Thai-language question answering system that gathers real-time traffic data from Twitter. The data collected are converted into traffic incident knowledge of what is happening and where it is happening. The system can then infer which points of interest (POIs) are affected by the incidents. Users can use natural (Thai) language to query the system against the ontology to receive traffic-related information. The system is currently deployed for demonstration on the web and developers can utilize it via REST API.