Research map / Label-efficient and robust learning / Recognition and long-tail
Spectral Word Embedding with Negative Sampling
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 |
|---|---|---|---|
| keep negative PMI values (alpha=-2.5) instead of PPMI (alpha=0) before SVD (ablation) | SVD-NS · Word analogy and word similarity | accuracy / correlation | Helped |
| use all negative PMI values (very low threshold) rather than a moderate threshold (ablation) | SVD-NS · Word analogy and word similarity | accuracy / correlation | Hurt |
| SVD-NS vs SGNS | SVD-NS · Google analogy | accuracy | Hurt |
| SVD-NS vs other embedding methods | SVD-NS · 8 word similarity datasets | correlation | Helped |
Closest papers by meaning
- Fast PMI-Based Word Embedding with Efficient Use of Unobserved Patterns (AAAI 2019 · similarity 0.89)
- Learning Semantic Representations for Novel Words: Leveraging Both Form and Context (AAAI 2019 · similarity 0.87)
- Distributed Negative Sampling for Word Embeddings (AAAI 2017 · similarity 0.87)
- On the Dimensionality of Word Embedding (NeurIPS 2018 · similarity 0.87)
- A Simple but Tough-to-Beat Baseline for Sentence Embeddings (ICLR 2017 · similarity 0.87)
- Using k-Way Co-Occurrences for Learning Word Embeddings (AAAI 2018 · similarity 0.87)
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.