Research map / Large language models / Retrieval and knowledge
A Pattern-Based Approach to Recognizing Time Expressions
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| PTime vs second-best system (SynTime) | PTime · WikiWars | strict F1: 80.11 -> 87.21 | Helped |
| PTime vs second-best system | PTime · WikiWars | relaxed F1: 94.25 -> 96.37 | Helped |
| PTime vs TOMN on Tweets | PTime · Tweets | strict F1: 92.56 -> 93.50 | Helped |
| PTime vs SynTime on Tweets | PTime · Tweets | relaxed F1: 95.87 -> 98.53 | Helped |
| PTime vs SynTime on TempEval-3 | PTime · TempEval-3 | strict F1: 92.09 -> 84.25 | Hurt |
| PTime vs TOMN on TempEval-3 | PTime · TempEval-3 | relaxed F1: 94.51 -> 91.58 | Hurt |
| manually add common patterns shared across WikiWars and Tweets (43 patterns, 14 not in TimeBank) (ablation) | PTime · TempEval-3 | strict F1: 84.25 -> 87.54 | Helped |
| pattern TIME UNIT dropped by cost bound due to negatives like 37 years old | PTime · TempEval-3 | recall | Hurt |
| set precision-control rho by grid search (0.87 / 0.94 / 0.94) | PTime · TempEval-3, WikiWars, Tweets | strict F1 | Helped |
| dynamically create one-token types for untyped words (e.g. hr) | PTime · Tweets | strict F1 | Helped |
| generate patterns with rich structure instead of heuristic boundary rules | PTime · WikiWars | strict recall: 80.22 -> 87.57 | Helped |
Closest papers by meaning
- A Semantic Parsing Framework for End-to-End Time Normalization (NeurIPS 2025 · similarity 0.84)
- TimE: A Multi-level Benchmark for Temporal Reasoning of LLMs in Real-World Scenarios (NeurIPS 2025 · similarity 0.80)
- Temporal Specification Optimisation for the Event Calculus (AAAI 2025 · similarity 0.80)
- NeSTR: A Neuro-Symbolic Abductive Framework for Temporal Reasoning in Large Language Models (AAAI 2026 · similarity 0.79)
- ITFormer: Bridging Time Series and Natural Language for Multi-Modal QA with Large-Scale Multitask Dataset (ICML 2025 · similarity 0.79)
- Test of Time: A Benchmark for Evaluating LLMs on Temporal Reasoning (ICLR 2025 · similarity 0.79)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.