Twitter Mining for Traffic Events Detection

Nowadays with the proliferation of smartphones and tablets on the market, almost everyone has access to mobile devices that offer better processing capabilities and access to new information and services. The Web is undoubtedly the best tool for sharing content, especially through social networks. One of the most useful information that can be extracted is the geographical one. Current navigation systems lack in several ways to satisfy the need to process and reason upon such volumes of data, namely, to accurately provide information about urban traffic in real-time and the possibility to personalize the information used by such systems. This paper describes an approach to integrate and fuse tweet messages from traffic agencies in UK, with the objective of detecting the geographical focus of traffic events. Tweet messages are considered in this work given its uniqueness, the real time nature of tweets which may be used to quickly detect a traffic event and its simplicity; it only cost 140 characters to generate a message (called "tweet") for any user. The approach presented here is composed by several steps: tweet classification, event type classification, name entity recognition, geolocation and event tracking. Finally, we do an experimental study on a real dataset composed by traffic related tweet messages to access the accuracy of proposed approach. We present some inaccuracies ranging from lack of geographical information, imprecise and ambiguous toponyms, overlaps and repetitions as well as visualization to our data set in UK. We finally give an outlook into potential corrections. The work presented here is still part of on-going work. Results achieved so far do not address the final conclusions but form the basis for the formalization of a domain knowledge along with the services.