Research map / Label-efficient and robust learning / Recognition and long-tail
Distributed Representations for Arithmetic Word Problems
Sowmya S Sundaram, Deepak P, Savitha Sam Abraham
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 |
|---|---|---|---|
| add GCN layer to Bi-LSTM (ablation) | Bi-LSTM Siamese · SingleOp schema retrieval | P@3: 0.7272 -> 0.8114 | Improved |
| add Bi-LSTM to GCN (ablation) | GCN Siamese · SingleOp schema retrieval | P@3: 0.7503 -> 0.8114 | Improved |
| add GCN layer to Bi-LSTM (ablation) | Bi-LSTM Siamese · SingleOp schema retrieval | NDCG@10: 0.8618 -> 0.8876 | Improved |
| operator-based augmentation with Siamese model vs LSTM autoencoder | lstm-ae · SingleOp schema retrieval | P@10: 0.6413 -> 0.7129 | Improved |
| add Bi-LSTM to GCN (ablation) | GCN Siamese · SingleOp classification AdaBoost | accuracy: 0.5664 -> 0.6726 | Improved |
| add GCN layer to Bi-LSTM (ablation) | Bi-LSTM Siamese · SingleOp classification MLP | accuracy: 0.7434 -> 0.8761 | Improved |
| add Bi-LSTM to GCN (ablation) | GCN Siamese · SingleOp classification Random Forest | accuracy: 0.8142 -> 0.8673 | Improved |
| use Ours instead of lstm-ae for random forest classification | lstm-ae · SingleOp classification | accuracy: 0.8407 -> 0.8673 | Improved |
| use Ours instead of lstm-dae for equation-based retrieval P@10 | lstm-dae · SingleOp equation retrieval | P@10: 0.4053 -> 0.5162 | Improved |
| use Ours instead of lstm-dae for equation-based retrieval NDCG@10 | lstm-dae · SingleOp equation retrieval | NDCG@10: 0.7284 -> 0.8032 | Improved |
| transfer SingleOp-trained model to MAWPS | Ours · MAWPS equation retrieval | P@10: 0.2498 -> 0.3015 | Improved |
| transfer to MAWPS | Ours · MAWPS equation retrieval | P@3: 0.6797 -> 0.4862 | Hurt |
Closest papers by meaning
- HMS: A Hierarchical Solver with Dependency-Enhanced Understanding for Math Word Problem (AAAI 2021 · similarity 0.87)
- A Compact Model for Mathematics Problem Representations Distilled from BERT (AAAI 2025 · similarity 0.86)
- MathNet: A Global Multimodal Benchmark for Mathematical Reasoning and Retrieval (ICLR 2026 · similarity 0.86)
- Unit Dependency Graph and Its Application to Arithmetic Word Problem Solving (AAAI 2017 · similarity 0.86)
- SMART: A Situation Model for Algebra Story Problems via Attributed Grammar (AAAI 2021 · similarity 0.85)
- Augmenting Math Word Problems via Iterative Question Composing (AAAI 2025 · similarity 0.84)
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.
