Research map / Large language models / Retrieval and knowledge
Exploring Knowledge Graphs in an Interpretable Composite Approach for Text Entailment
Vivian S. Silva, André Freitas, Siegfried Handschuh
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 |
|---|---|---|---|
| use composite routing of syntactic and semantic models instead of syntactic EditDistance (ablation) | entailment system · BPI | F-measure: 0.53 -> 0.66 | Improved |
| use composite model instead of semantic-only GraphNavigation (ablation) | entailment system · BPI | F-measure: 0.59 -> 0.66 | Improved |
| use composite model instead of GraphNavigation (ablation) | entailment system · BPI | precision: 0.65 -> 0.57 | Hurt |
| use composite model instead of EditDistance (ablation) | entailment system · GHS | F-measure: 0.45 -> 0.60 | Improved |
| use composite model instead of GraphNavigation (ablation) | entailment system · GHS | F-measure: 0.53 -> 0.60 | Improved |
| use composite model vs MaxEntClassifier | entailment system · GHS | F-measure: 0.66 -> 0.60 | Hurt |
| use Wiktionary graph instead of WordNet (ablation) | CompositeEntailment · BPI | F-measure: 0.66 -> 0.60 | Hurt |
| use Wikipedia graph instead of WordNet (ablation) | CompositeEntailment · BPI | F-measure: 0.66 -> 0.32 | Hurt |
| use Wikipedia graph instead of WordNet (ablation) | CompositeEntailment · GHS | recall: 0.70 -> 0.11 | Hurt |
| use Wiktionary graph instead of WordNet (ablation) | CompositeEntailment · GHS | F-measure: 0.60 -> 0.58 | Hurt |
| larger knowledge graph (Wikipedia, 859,087 defs) vs smaller WordNet (93,699) (ablation) | CompositeEntailment · BPI | recall: 0.79 -> 0.26 | Hurt |
Closest papers by meaning
- Recognizing and Justifying Text Entailment Through Distributional Navigation on Definition Graphs (AAAI 2018 · similarity 0.95)
- SciTaiL: A Textual Entailment Dataset from Science Question Answering (AAAI 2018 · similarity 0.87)
- Infusing Knowledge into the Textual Entailment Task Using Graph Convolutional Networks (AAAI 2020 · similarity 0.87)
- Defeasible Visual Entailment: Benchmark, Evaluator, and Reward-Driven Optimization (AAAI 2025 · similarity 0.83)
- Graph Reasoning Transformers for Knowledge-Aware Question Answering (AAAI 2024 · similarity 0.83)
- Harnessing Large Language Models for Knowledge Graph Question Answering via Adaptive Multi-Aspect Retrieval-Augmentation (AAAI 2025 · similarity 0.83)
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.
