Research map / Vision-language and multimodal / Human and embodied modeling
LET: Linguistic Knowledge Enhanced Graph Transformer for Chinese Short Text Matching
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 |
|---|---|---|---|
| Add the LET lattice-and-sense matching approach to BERT | BERT · BQ test | accuracy: 84.50 -> 85.30 | Helped |
| Add the LET lattice-and-sense matching approach to BERT | BERT · LCQMC test | accuracy: 85.73 -> 88.38 | Helped |
| Use a word lattice instead of one segmentation path (ablation) | LET-BERT · LCQMC test | accuracy: 87.84 -> 88.38; 87.72 -> 88.38; 87.50 -> 88.38 | Helped |
| Add HowNet sense information and sense updating (ablation) | LET-BERT with lattice input · LCQMC test | accuracy: 87.68 -> 88.38 | Helped |
| Add HowNet sense information and sense updating (ablation) | LET-BERT with lattice input · LCQMC test | F1: 88.40 -> 88.85 | Helped |
| Add the GRU to SaGT to gate contextual and semantic messages (ablation) | LET-BERT · LCQMC test | accuracy: 87.82% | Helped |
| Use two SaGT layers (ablation) | LET-BERT · LCQMC and BQ | accuracy | Helped |
| Add HowNet sense information (ablation) | LET-BERT · LCQMC test, text length <16 | accuracy: 88.99 -> 90.05 | Helped |
| Add HowNet sense information (ablation) | LET-BERT · LCQMC test, text length >22 | accuracy: 84.53 -> 85.13 | Helped |
Closest papers by meaning
- Lattice CNNs for Matching Based Chinese Question Answering (AAAI 2019 · similarity 0.88)
- LADA-Trans-NER: Adaptive Efficient Transformer for Chinese Named Entity Recognition Using Lexicon-Attention and Data-Augmentation (AAAI 2023 · similarity 0.87)
- A Systematic Exploration of Knowledge Graph Alignment with Large Language Models in Retrieval Augmented Generation (AAAI 2025 · similarity 0.84)
- Multi-View Empowered Structural Graph Wordification for Language Models (AAAI 2025 · similarity 0.83)
- Talk like a Graph: Encoding Graphs for Large Language Models (ICLR 2024 · similarity 0.83)
- RMIB: Representation Matching Information Bottleneck for Matching Text Representations (ICML 2024 · 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.