Research map / Label-efficient and robust learning / Recognition and long-tail
cw2vec: Learning Chinese Word Embeddings with Stroke n-gram Information
Shaosheng Cao, Wei Lu, Jun Zhou, Xiaolong Li
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 |
|---|---|---|---|
| represent words by stroke n-grams (cw2vec) vs component-based JWE | Chinese word embeddings · word analogy 3CosAdd | accuracy: 74.2 -> 78.1 | Improved |
| cw2vec vs JWE | Chinese word embeddings · BosonNLP NER | F1: 67.9 -> 71.7 | Improved |
| cw2vec vs best baseline | Chinese word embeddings · Fudan text classification | accuracy: 94.3 -> 95.3 | Improved |
| cw2vec vs JWE | Chinese word embeddings · wordsim-296 | Spearman rho x100: 52.7 -> 52.7 | No clear effect |
| cw2vec vs CWE | Chinese word embeddings · wordsim-240 | Spearman rho x100: 50.0 -> 50.4 | Improved |
| add character information (CWE vs skip-gram) | word2vec · word analogy 3CosAdd | accuracy: 58.3 -> 68.5 | Improved |
| use pixel-based glyph features (GWE vs skip-gram) | word2vec · word analogy 3CosAdd | accuracy: 58.3 -> 50.8 | Hurt |
Closest papers by meaning
- What the Vec? Towards Probabilistically Grounded Embeddings (NeurIPS 2019 · similarity 0.86)
- Closed-Form Training Dynamics Reveal Learned Features and Linear Structure in Word2Vec-like Models (NeurIPS 2025 · similarity 0.85)
- Semantic Structure-Based Word Embedding by Incorporating Concept Convergence and Word Divergence (AAAI 2018 · similarity 0.85)
- Improving Word Embeddings with Convolutional Feature Learning and Subword Information (AAAI 2017 · similarity 0.85)
- SSMI: Semantic Similarity and Mutual Information Maximization Based Enhancement for Chinese NER (AAAI 2023 · similarity 0.85)
- Repurposing Language Models into Embedding Models: Finding the Compute-Optimal Recipe (NeurIPS 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.
