Improving Road Traffic Safety by Mining Accident

Data: Developing a Decision Tree to Classify Injury

Severity Using R and Hadoop
In the developed as well as developing countries, infrastructure development is one of the major
investment by the government, while the safety of the passengers on roads is of utmost
importance, the budgetary constraints leave a void on the quality of the roads. A road
optimization during the construction or during the maintenance phase, requires that the engineers
analyse all the parameters that play a crucial role in ensuring safety for the passengers and
preventing accidents. The data to be analysed is collected from various sources, is both
structured and unstructured (raw) and has several attributes. It is a challenge to gather all such
relevant data, detect and analyse it together to generate decision trees that give insights on
previous accidents. For this purpose, we propose to harness the power of Big Data technologies
like Hadoop Map Reduce and process tools like R. The analysis will be represented in the form
of a decision tree which can be represented graphically.