Research map / Label-efficient and robust learning / Recognition and long-tail
Representations of Context in Recognizing the Figurative and Literal Usages of Idioms
Changsheng Liu, Rebecca Hwa
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| weighted late fusion of lexical, topical and distributional representations vs best competing method | idiom usage recognizer · SemEval 2013 5B, bread and butter / break a leg / drop the ball | Ffig: 0.61 -> 0.72 (drop the ball); 0.83 -> 0.81 (through the roof) | Mixed |
| late fusion narrows variance across idioms | idiom usage recognizer · SemEval 2013 5B | Ffig range: 0.68 to 0.85 range vs 0.54 to 0.83 for Rajani et al. | Improved |
| learn adaptive weights (late fusion) instead of concatenating features (early fusion) (ablation) | fusion model · SemEval 2013 5B, bread and butter | Ffig: 0.82 -> 0.84 | Improved |
| learn adaptive weights (late fusion) instead of concatenating features (early fusion) (ablation) | fusion model · SemEval 2013 5B, in the fast lane | Ffig: 0.54 -> 0.72 | Improved |
| combine all three representations rather than distributional only (ablation) | late fusion model · SemEval 2013 5B, bread and butter | Ffig: 0.57 -> 0.84 | Improved |
| combine all three representations rather than lexical only (ablation) | late fusion model · SemEval 2013 5B, break a leg | Ffig: 0.58 -> 0.73 | Improved |
| remove additional non-contextual features from supervised baselines (ablation) | Rajani et al. · SemEval 2013 5B | Ffig: 0.54 -> 0.57 (break a leg) | No clear effect |
| higher context diversity (more LDA topics) for lexical/topical methods | idiom recognition · SemEval 2013 5B | Ffig: r = -0.86 | Hurt |
| higher semantic analyzability (similarity of literal and figurative contexts) for distributional method | Sporleder and Li · SemEval 2013 5B | Ffig: r = -0.77 | Hurt |
Closest papers by meaning
- A Generalized Idiom Usage Recognition Model Based on Semantic Compatibility (AAAI 2019 · similarity 0.94)
- Translate Meanings, Not Just Words: IdiomKB’s Role in Optimizing Idiomatic Translation with Language Models (AAAI 2024 · similarity 0.85)
- Mitigating Idiom Inconsistency: A Multi-Semantic Contrastive Learning Method for Chinese Idiom Reading Comprehension (AAAI 2024 · similarity 0.84)
- Idiomatic Expression Paraphrasing without Strong Supervision (AAAI 2022 · similarity 0.84)
- McHirc: A Multimodal Benchmark for Chinese Idiom Reading Comprehension (AAAI 2025 · similarity 0.83)
- ImpScore: A Learnable Metric For Quantifying The Implicitness Level of Sentences (ICLR 2025 · similarity 0.81)
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.
