Knowledge Engineering with Semantic Technologies to Identify and Warn of Transport Disruptions

Tracking #: 2640-3854

This paper is currently under review
David Corsar1
Milan Markovic
Peter Edwards
Paul Gault
Caitlin Cottrill
John D. Nelson
Somayajulu Sripada

Responsible editor: 
Guest Editors Transportation Data 2020

Submission type: 
Full Paper
Public transport operators and transport authorities are increasingly using social media channels to disseminate relevant information to travellers. Many operators utilise platforms such as Twitter to provide both customer service and real-time passenger information - including details about service disruptions. However, for such reports to be useful to travellers, they must first find the information, relate it to their travel plans to determine if their journey will be adversely impacted, and, if so, decide how to adapt their plans. This paper describes the TravelBot intelligent system, developed to perform the first two of these tasks and warn public transport users of potential disruptions to their journey. Developing TravelBot combined knowledge engineering processes with iterative user-led design activities culminating in a real-world user evaluation. Semantic Web technologies are used to represent and integrate transport knowledge obtained from open data, social media posts, and users, and to support reasoning processes that infer structured representations of events described in social media posts. Inferred events are assessed to determine if they are are likely to disrupt the planned travel of TravelBot users, and, if so, users are sent personalised warnings. Evaluations of the system based on data collected during a user trial found that social media posts are processed in an acceptable length of time, and users generally considered the information provided by the system to be useful. Areas in which the event inference capability could be improved are also identified and offer future research opportunities.
Full PDF Version: 
Under Review