In the transportation field, a huge amount of information has been gathered by IoT devices, remote detecting and other information assortment apparatuses bring new challenges, the size of this information turns out to be amazingly large and increasingly complex for conventional methods of data mining. To manage this test, Apache Spark remain as an incredible huge scale disseminated registering stage that can be utilized effectively for AI against exceptionally huge databases. This work utilized enormous scale AI methods particularly Decision Tree with Apache Spark structure for large information investigation to fabricate a model that can foresee the elements lead to street mishaps dependent on a few information factors identified with car crashes. In light of this, the anticipating model first pre-forms the large mishap information and dissect it to make information for a learning framework. Observational outcomes show that the proposed model could give new data that can help the leaders to examine and improve street safety.

Keywords: transportation; IoT; data mining