UNIETD – Assessment of Third Party Data as Information Source for Drivers and Road Operators

The paper deals with the assessment of third party data such as crowd sourced/social media and floating vehicle data as information source for road operators in addition to traditional infrastructure-based techniques. For purposes of quality assessment of different types of data and available ground truths existing test/evaluation methodologies have been assessed. A new methodology has been designed for assessment of speeds and travel times using normalized (between 0 and 1) quality indicators that can distinguish between “detection rate” and “false alarm rate” concepts. In terms of harvesting social media the relevance of social media content has been assessed against a range of traffic management requirements. Furthermore the level of content that will be available has been estimated as well as commercial sources and business models for road authorities. Analyses cover unstructured data from Twitter and Facebook both historical data and three months of contemporary data. In addition surveys are conducted in England and Austria to retrieve information from the public in terms of which social media platforms are commonly used to share information about traffic related incidents.