This chapter considers the use of publicly available social media data as a potential additional source of traffic information. Social media data with geographical information may be useful for estimating the speed of traffic. Information on traffic flow, delays, infrastructure and environment -related traffic issues may be obtained from studying the textual content of the messages. This chapter is concerned with assessing the relevance of these social media data to the needs of road administrations, particularly in the context of traffic management. We aim to focus on the potential of one commonly available type of social media data, Twitter, as a new source of travel time information. We consider the efficacy of the data, its availability and different business models for accessing and processing the data. A case study is used to provide detailed illustration of some of the issues with the functional contribution of Twitter data and the surrounding eco-system.

Chapter Contents:

• 11.1 Introduction
• 11.2 Background and context
• 11.3 Twitter as a source of transport system data
• 11.3.1 Accessing Twitter data
• 11.3.2 A case study of Twitter for TTI
• 11.3.3 Twitter data set characteristics: time and location distribution
• 11.4 Travel time estimation from Twitter data
• 11.4.1 Calculation of Twitter-based speed
• 11.4.2 Correlation with loop speed data
• 11.5 Analysis of Twitter content for traffic management relevance
• 11.5.1 Definition of relevant content
• 11.5.2 Understanding relevance through stated traffic management objectives
• 11.5.3 Understanding relevance through ontology
• 11.5.4 Classifying tweets as relevant or not-relevant
• 11.5.5 Classifying tweets using ontology
• 11.6 Conclusions
• Acknowledgements
• References

Inspec keywords: social networking (online); geographic information systems; traffic information systems

Other keywords: business models; geographical information; travel time information; social media data; Twitter; traffic information

Subjects: Geography and cartography computing; Information networks; Traffic engineering computing