Negative Sampling

Negative Sampling is a training strategy in machine learning that belongs to the class of sampling techniques and is distinguished by its selection of a small number of contrasting (negative) examples—typically from a large set of possible non-observed interactions, entity pairs, or document passages—to improve computational efficiency and model discriminability in tasks such as relation extraction, passage retrieval, and implicit feedback modeling.

279

Publications

47.9K

Citations

941

Authors

283

Institutions

Publications per year

2017–2026

247

Authors

941

  1. SM
    Shaoping Ma

    Tsinghua University, China

    9 publications · 1K citations

    9

    h-index

  2. YL
    Yiqun Liu

    Tsinghua University, China

    9 publications · 1K citations

    9

    h-index

  3. JT
    Jie Tang

    Tsinghua University, China

    6 publications · 1.1K citations

    6

    h-index

  4. WM
    Weizhi Ma

    Tsinghua University, China

    5 publications · 535 citations

    5

    h-index

  5. MZ
    Min Zhang

    Tsinghua University, China

    5 publications · 667 citations

    5

    h-index

1–5 of 941

Venues

  1. arXiv (Cornell University)

    Repository · 64 publications · 31.6K citations

    27

    h-index

  2. Proceedings of the AAAI Conference on Artificial Intelligence

    Conference · 14 publications · 783 citations

    13

    h-index

  3. IEEE Transactions on Knowledge and Data Engineering

    TKDE · 6 publications · 278 citations

    6

    h-index

  4. Neurocomputing

    4 publications · 97 citations

    4

    h-index

  5. Proceedings of the ACM Web Conference 2022

    Conference · 4 publications · 115 citations

    4

    h-index

1–5