Researchers have developed a new method for computing deadlines from legal documents using temporal dependency graphs, which they found to be more accurate than direct answers from language models. The system correctly reproduces six out of seven timeliness rulings on UK Employment Appeal Tribunal judgments. While language models struggled with arithmetic and often contradicted their own reasoning, the graph-based approach achieved 90.2% accuracy on generated cases, with errors primarily stemming from the initial extraction of triggering events rather than the calculation itself. AI
IMPACT This research demonstrates a more reliable method for AI to process legal documents, potentially improving accuracy in time-sensitive legal computations.
RANK_REASON Academic paper detailing a new methodology for AI-assisted legal document analysis. [lever_c_demoted from research: ic=1 ai=1.0]
- alphaXiv
- arXiv
- CatalyzeX
- CORE Recommender
- DagsHub
- Gotit.pub
- Hugging Face
- Influence Flower
- ScienceCast
- UK Employment Appeal Tribunal
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
