Traffic
Contents
Traffic Management
Traffic management is a challenging issue in most of the major cities around the world. With increasing number of people moving to cities for economic opportunities, this trend is going to make city resource management a crucial problem. We take this important problem and study its characteristics for creating solutions toward a grand vision of providing actionable information to policy makers in a city.
Problem of Understanding City Traffic Events
A city is a complex system of systems with many heterogeneous components. With decreasing cost of sensors to monitor our environment, there are efforts to deploy sensors to monitor vehicular traffic. 511.org is an exemplar of such an effort which monitors traffic flow in San Francisco Bay Area. These sensors monitor vehicular speed and volume though various road links. These observations are then utilized to derive travel time for each links. The idea of deploying these sensors is to capture real-world events that exists in a city such as accidents, breakdowns, bad weather, etc. There are many semantic type for possible events in a city but there is a single mode of observation in this context which is the sensor observations.
City as a Physical-Cyber-Social System
A city is not limited to physical sensors
Solution
Event Extraction
Extracting City Traffic Events from Social Streams
Cities are composed of complex systems with physical, cyber, and social components. Current work on extracting and understanding city events mainly rely on technology enabled infrastructure to observe and record events. In this work, we propose an approach to leverage citizen observations of various city systems and services such as traffic, public transport, water supply, weather, sewage, and public safety as a source of city events. We investigate the feasibility of using such textual streams for extracting city events from annotated text. We formalize the problem of annotating social streams such as microblogs as a sequence labeling problem. We present a novel training data creation process for training sequence labeling models. Our automatic training data creation process utilizes instance level domain knowledge (e.g., locations in a city, possible event terms). We compare this automated process to a state-of-the-art tool that needs manually created training data and show that it has comparable performance in annotation tasks. An aggregation algorithm is then presented for event extraction from annotated text. We carry out a comprehensive evaluation of the event annotation and event extraction on a real-world dataset consisting of event reports and tweets collected for over four months from San Francisco Bay Area. The evaluation results are promising and provides insights into the utility of social stream for city events.
We use Open Science Framework to share project resources with the research community. Here is the link for our project on Open Science Framework.
