(PDF) From Prediction to Prevention: Using Text Mining ...

Executive Summary

This study develops a machine learning-based analytical framework to identify, quantify, and interpret the factors associated with severe bus accidents. The framework utilizes text mining and explainable machine learning techniques to analyze unstructured data from accident reports. By transforming this data into actionable insights, the study aims to move from merely predicting accidents to preventing them. The approach highlights the potential of AI in enhancing safety measures and reducing accident rates by providing a deeper understanding of the underlying causes of accidents. This research is particularly relevant for the insurance industry, where understanding risk factors is crucial for underwriting and claims processing.