Publications / 2018 Proceedings of the 35th ISARC, Berlin, Germany
Retrieving proper accident cases and extracting risk factors from them are crucial for construction safety management. However, the process was often challenging due to unstructured properties of text data in accident reports, which caused limited, inefficient, and non-consistent information retrieval and knowledge gathering. To overcome the problems, this research aimed at developing a semantic search system to retrieve proper accident cases based on users deliberate intentions and to extract safety risk factors automatically using Natural Language Processing. The performance of the system prototype was evaluated by construction practitioners with promising results for more usable construction accident database development (i.e., thesaurus) and efficient accident analysis using the thesaurus.