Research map / Label-efficient and robust learning / Recognition and long-tail
Multi-Label Patent Categorization with Non-Local Attention-Based Graph Convolutional Network
Pingjie Tang, Meng Jiang, Bryan (Ning) Xia, Jed W. Pitera, Jeffrey Welser, Nitesh V. Chawla
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 |
|---|---|---|---|
| replace GraphSAGE with attention-based BiLSTM aggregator (A-GCN) (ablation) | GraphSAGE · Patent CIRCA 45k | P@1: 72.95 -> 75.74 | Improved |
| add AttentionXML-style label attention to A-GCN (ablation) | A-GCN · Patent CIRCA 45k | P@1: 75.74 -> 76.72 | Improved |
| replace label attention with adaptive non-local second-order attention (ablation) | A-GCN+LA · Patent CIRCA 45k | P@1: 76.72 -> 78.30 | Improved |
| replace label attention with adaptive non-local second-order attention (ablation) | A-GCN+LA · Patent CIRCA 45k | nDCG@5: 77.20 -> 78.28 | Improved |
| add adaptive non-local second-order attention to A-GCN (ablation) | A-GCN · Patent CIRCA 45k | P@5: 34.92 -> 36.41 | Improved |
| add phrase features to bag-of-words (ablation) | Parabel · Patent CIRCA 45k | P@1: 75.10 -> 75.79 | Improved |
| add phrase features to bag-of-words (ablation) | FastXML · Patent CIRCA 45k | P@1: 73.10 -> 73.92 | Improved |
| use 20 attention heads (ablation) | A-GCN+A-NLSOA · Patent CIRCA 45k | P@1 | Improved |
| use sliding window size 20 for PMI instead of 5, 10 or 25 (ablation) | A-GCN+A-NLSOA · Patent CIRCA 45k | P@1 | Improved |
| use covariance without power normalization (ablation) | A-NLSOA · Patent CIRCA 45k | P@1 | Improved |
| use softmax rather than sigmoid output (ablation) | A-GCN+A-NLSOA · Patent CIRCA 45k | P@1 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- AttentionXML — AttentionXML: Label Tree-based Attention-Aware Deep Model for High-Performance Extreme Multi-Label Text Classification (NeurIPS 2019) · better on 6 of 6 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- PatentLMM: Large Multimodal Model for Generating Descriptions for Patent Figures (AAAI 2025 · similarity 0.83)
- The Harvard USPTO Patent Dataset: A Large-Scale, Well-Structured, and Multi-Purpose Corpus of Patent Applications (NeurIPS 2023 · similarity 0.83)
- All-in Text: Learning Document, Label, and Word Representations Jointly (AAAI 2016 · similarity 0.83)
- IMPACT: A Large-scale Integrated Multimodal Patent Analysis and Creation Dataset for Design Patents (NeurIPS 2024 · similarity 0.82)
- TGLsta: Low-resource Textual Graph Learning with Semantic and Topological Awareness via LLMs (AAAI 2025 · similarity 0.81)
- Don't Pay Attention, PLANT It: Pretraining Attention via Learning-to-Rank (NeurIPS 2026 · 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.
