Search
  • Soheil Sohrabi

Traffic Safety and Social Media

Updated: Nov 12, 2018


The effectiveness of traditional traffic incident data collection is limited by sparse sensor coverage as well as cost and labor intensive processes. These shortcomings in crash data collection coupled with emerging technologies have incentivized researchers to propose innovative data collecting techniques based on crowed-sourcing data. In this context, Twitter is widely used in the literature to extract the traffic incident information. Twitter currently produces 350,000 tweets per minute from more than 335 million monthly active users, which represent the potential for extracting information from individual or public tweets. Previous work has shown that about 80% of incidents can be found in Twitter databases. So far, very little research has focused on extracting data specifically related to crashes involving vehicles and vulnerable road users (e.g., bikers and walkers).


Vulnerable users crashes are not reporting properly and the crash information are not fully represent the crash frequency and information. by exploring Tweets the accidents related to vulnerable users can be detected and required information such as time of the crash, emergency services and crash reason can be extracted. The findings can be used for detecting the crash hot-spots, designing more efficient active transportation friendly infrastructures, imposing policies to reduce the crash severity such as reducing the traffic speed and improving more efficient biking laws.


What else do you think the Tweet data can contribute to traffic safety?



12 views0 comments