NSW Traffic Watch

- Utilizing the crow sourcing real-time information, TrafficWatch is a system that leverages Twitter as a channel for transport network reporting, incident and event managements.
- The system is supported by advanced natural languages processing technologies and state-of-the-art machine learning algorithms (Support Vector Machines and Conditional Random Fields).
- As Twitter API provides data from all over the world, the crawled tweets are filtered to get incidents from Australia only.
- The tweets contain key-words like "accident" or "crash", etc. but are not related to traffic are also classified and removed.
- Incident locations are extracted from the text and visualized on the map.
- Any travelers can easily contribute real-time information to the system.
- If you are on the bus, train, passenger or pedestrian (not driving please) and you see the accident, simply tweet about it with some details about the locations (or just turn on your geo-location), we will get it and display in the system within seconds.
- The more people tweet about an accident, the more reliability it should be.
- Findings from the use of TrafficWatch at Data61 demonstrated that it has potential to report incidents earlier than other data sources, as well as identifying unreported incidents.

*Not many apps have the advances of both natural languages processing, machine learning and map visualisation like this one.
*Free app with no ads.