Reproduce “HSEarch: semantic search system for workplace accident reports” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Emrah Inan, Paul Thompson, Tim Yates, Sophia Ananiadou
HSEarch is a semantic search system developed for workplace accident reports in the construction industry, combining keyword search with advanced text mining techniques to improve information retrieval. It utilizes Named Entity Recognition, document clustering, and summarization to enable interactive exploration and more precise access to safety information. The system achieves an nDCG of 0.937 for entity-based indexing, outperforming traditional word-based methods.
Key insights
- Integrating domain-specific Named Entity Recognition into the search indexing process significantly enhances the quality of ranked retrieval results for specialized document collections.
- Combining multiple text mining techniques, including NER, clustering, and summarization, within an interactive interface provides a powerful and multifaceted approach to exploring complex, unstructured textual data.
- Semantic search capabilities, such as faceted filtering by identified entities and dynamic document clustering, enable experts to extract actionable knowledge more efficiently and flexibly from incident reports.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.