Browse › Natural Language Processing › Named Entity Recognition (NER) › CoNLL 2003 (English)
CoNLL 2003 (English) Benchmark (Named Entity Recognition (NER))
Named Entity Recognition (NER) is a task of Natural Language Processing (NLP) that involves identifying and classifying named entities in a text into predefined categories such as person names, organizations, locations, and others. The goal of NER is to extract structured information from unstructured text data and represent it in a machine-readable format. Approaches typically use BIO notation, which differentiates the beginning (B) and the inside (I) of entities. O is used for non-entity tokens.
Example:
| Mark | Watney | visited | Mars |
|---|---|---|---|
| B-PER | I-PER | O | B-LOC |
The archive carries no text for this table; the description above is the archive's text for the task Named Entity Recognition (NER). archive 2025-07-28
Over time archive 2025-07-28
The chart needs JavaScript; the table below carries every value.
Direction inferred from the metric name, not from the archive: F1 (higher is better). Points are placed at the row's paper date; 72 of 73 rows carry one.
Results archive 2025-07-28
Archive rows end at the archive snapshot, 2025-07-28: no result published after that date is in this table. Rank is the archive's row order at that snapshot; not re-ranked here. Metric values are the archive's strings. Column headers sort the table in your browser; each row keeps its archive rank.
| Paper | Code | Ran Syntology | Report | |||||
|---|---|---|---|---|---|---|---|---|
| 1 | ACE + document-context | 94.6 | – | Paper | Code | 2020 | linked, not harvested | report |
| 2 | LUKE 483M | 94.3 | – | Paper | Code | 2020 | 3 of 10 ran · 7 unverified | report |
| 3 | Co-regularized LUKE | 94.22 | – | Paper | Code | 2021 | linked, not harvested | report |
| 4 | LUKE + SubRegWeigh (K-means) | 94.2 | – | Paper | Code | 2024 | linked, not harvested | report |
| 5 | ASP+T5-3B | 94.1 | – | Paper | Code | 2022 | linked, not harvested | report |
| 6 | FLERT XLM-R | 94.09 | ✓ | Paper | Code | 2020 | linked, not harvested | report |
| 7 | PL-Marker | 94.0 | – | Paper | Code | 2021 | 3 of 9 ran · 6 unverified | report |
| 8 | CL-KL | 93.85 | – | Paper | Code | 2021 | 0 of 4 ran · 4 unverified | report |
| 9 | XLNet-GCN | 93.82 | – | Paper | Code | 2021 | linked, not harvested | report |
| 10 | RoBERTa + SubRegWeigh (K-means) | 93.81 | – | Paper | Code | 2024 | linked, not harvested | report |
| 11 | ASP+flan-T5-large | 93.8 | – | Paper | Code | 2022 | linked, not harvested | report |
| 12 | InferNER | 93.76 | – | Paper | – | 2021 | no code linked | report |
| 13 | Cross-sentence context (First) | 93.74 | – | Paper | Code | 2020 | 1 of 7 ran · 6 unverified | report |
| 14 | XLM-RoBERTa-large union | 93.69 | – | Paper | – | 2024 | no code linked | report |
| 15 | Baseline + BS | 93.65 | – | Paper | Code | 2022 | linked, not harvested | report |
| 16 | ACE | 93.64 | – | Paper | Code | 2020 | linked, not harvested | report |
| 17 | BERT-CRF | 93.6 | – | Paper | Code | 2021 | linked, not harvested | report |
| 18 | CNN Large + fine-tune | 93.5 | ✓ | Paper | – | 2019 | no code linked | report |
| 19 | Biaffine-NER | 93.5 | – | Paper | Code | 2020 | 3 of 10 ran · 7 unverified | report |
| 20 | GCDT + BERT-L | 93.47 | ✓ | Paper | Code | 2019 | linked, not harvested | report |
| 21 | I-DARTS + Flair | 93.47 | ✓ | Paper | Code | 2019 | linked, not harvested | report |
| 22 | CrossWeigh + Pooled Flair | 93.43 | – | Paper | Code | 2019 | 4 of 4 ran · 0 unverified | report |
| 23 | LSTM-CRF+ELMo+BERT+Flair | 93.38 | – | Paper | Code | 2019 | linked, not harvested | report |
| 24 | Hierarchical + BERT | 93.37 | – | Paper | Code | 2019 | 1 of 9 ran · 8 unverified | report |
| 25 | BERT-CRF (Replicated in AdaSeq) | 93.35 | – | Paper | Code | 2021 | 0 of 4 ran · 4 unverified | report |
| 26 | BERT-MRC+DSC | 93.33 | ✓ | Paper | Code | 2019 | 0 of 3 ran · 3 unverified | report |
| 27 | XLNet | 93.28 | – | Paper | Code | 2021 | linked, not harvested | report |
| 28 | BARTNER | 93.24 | – | Paper | Code | 2021 | linked, not harvested | report |
| 29 | GoLLIE | 93.1 | ✓ | Paper | Code | 2023 | 18 of 25 ran · 7 unverified | report |
| 30 | Flair embeddings | 93.09 | ✓ | Paper | Code | 2018 | linked, not harvested | report |
| 31 | PromptNER [RoBERTa-large] | 93.08 | ✓ | Paper | Code | 2023 | linked, not harvested | report |
| 32 | W2NER | 93.07 | – | Paper | Code | 2021 | linked, not harvested | report |
| 33 | BERT-MRC | 93.04 | ✓ | Paper | Code | 2019 | 2 of 5 ran · 3 unverified | report |
| 34 | Locate and Label | 92.94 | – | Paper | Code | 2021 | linked, not harvested | report |
| 35 | PIQN | 92.87 | – | Paper | Code | 2022 | linked, not harvested | report |
| 36 | DiffusionNER | 92.78 | – | Paper | Code | 2023 | linked, not harvested | report |
| 37 | HSCRF + softdict | 92.75 | ✓ | Paper | Code | 2019 | linked, not harvested | report |
| 38 | TENER | 92.62 | – | Paper | Code | 2019 | 0 of 1 ran · 1 unverified | report |
| 39 | CVT + Multi-Task | 92.61 | ✓ | Paper | Code | 2018 | linked, not harvested | report |
| 40 | CVT + Multi-Task + Large | 92.61 | – | Paper | Code | 2018 | linked, not harvested | report |
| 41 | Stack LSTM | 92.43 | – | Paper | – | 2019 | no code linked | report |
| 42 | PromptNER [BERT-large] | 92.41 | – | Paper | Code | 2023 | linked, not harvested | report |
| 43 | DGLSTM-CRF + ELMo (L=2) 3.0pt1-4.51.5 | 92.4 | – | Paper | Code | 2019 | linked, not harvested | report |
| 44 | GRN | 92.34 | – | Paper | Code | 2019 | linked, not harvested | report |
| 45 | Neural-CRF+AE | 92.29 | – | Paper | Code | 2018 | 6 of 6 ran · 0 unverified | report |
| 46 | MGNER | 92.28 | – | Paper | Code | 2019 | linked, not harvested | report |
| 47 | BiLSTM-CRF+ELMo | 92.22 | ✓ | Paper | Code | 2018 | 23 of 58 ran · 35 unverified | report |
| 48 | SpanRel | 92.2 | – | Paper | Code | 2019 | linked, not harvested | report |
| 49 | LD-Net | 92.03 | – | Paper | Code | 2018 | linked, not harvested | report |
| 50 | GCDT | 91.96 | – | Paper | Code | 2019 | linked, not harvested | report |
| 51 | Hierarchical | 91.96 | – | Paper | Code | 2019 | 1 of 9 ran · 8 unverified | report |
| 52 | CRF + AutoEncoder | 91.87 | – | Paper | Code | 2018 | 6 of 6 ran · 0 unverified | report |
| 53 | PRISM | 91.8 | – | Paper | Code | 2019 | linked, not harvested | report |
| 54 | GraphIE (GCN+BiLSTM) | 91.74 | – | – | – | not matched | report | |
| 55 | Bi-LSTM-CRF + Lexical Features | 91.73 | – | Paper | Code | 2018 | linked, not harvested | report |
| 56 | IntNet + BiLSTM-CRF | 91.64 | – | Paper | – | 2018 | no code linked | report |
| 57 | Yang et al. ([2017a]) | 91.62 | – | Paper | Code | 2017 | linked, not harvested | report |
| 58 | Bi-LSTM-CNN | 91.62 | – | Paper | Code | 2015 | 4 of 9 ran · 5 unverified | report |
| 59 | S-LSTM | 91.57 | – | Paper | Code | 2018 | 1 of 1 ran · 0 unverified | report |
| 60 | LSTM with dynamic skip | 91.56 | – | Paper | Code | 2018 | linked, not harvested | report |
| 61 | Adversarial Bi-LSTM | 91.56 | – | Paper | Code | 2017 | linked, not harvested | report |
| 62 | HSCRF | 91.38 | – | Paper | Code | 2018 | linked, not harvested | report |
| 63 | IXA pipes | 91.36 | – | Paper | Code | 2017 | linked, not harvested | report |
| 64 | NCRF++ | 91.35 | – | Paper | Code | 2018 | 1 of 9 ran · 8 unverified | report |
| 65 | Yang et al. | 91.26 | – | Paper | Code | 2017 | 0 of 8 ran · 8 unverified | report |
| 66 | LM-LSTM-CRF | 91.24 | – | Paper | Code | 2017 | linked, not harvested | report |
| 67 | Bi-LSTM-CNN-CRF | 91.22 | – | Paper | Code | 2018 | linked, not harvested | report |
| 68 | BLSTM-CNN-CRF | 91.21 | – | Paper | Code | 2016 | 4 of 24 ran · 20 unverified | report |
| 69 | Bi-LSTM + Logic rules | 91.18 | – | Paper | Code | 2016 | linked, not harvested | report |
| 70 | LSTM-CRF | 90.94 | – | Paper | Code | 2016 | 10 of 36 ran · 26 unverified | report |
| 71 | GCN | 88.63 | – | Paper | Code | 2021 | linked, not harvested | report |
| 72 | SWEM-CRF | 86.28 | – | Paper | Code | 2018 | linked, not harvested | report |
| 73 | VSL-GG-Hier | 84.7 | – | Paper | Code | 2019 | 1 of 1 ran · 0 unverified | report |
All 73 rows shown. 72 link to a paper page on this site; 12 are marked as using additional training data in the archive. No GitHub stars are tracked; "Code" is the first repository the archive lists for the row. The archive carries no row tags, review links or community-submitted rows for this table; none are shown. archive 2025-07-28
Syntology Ran reads "N of M ran · U unverified": of the M code samples Syntology harvested from repositories linked to that row's paper (joined by arXiv id), N executed on a synthesized input and the other U = M−N are unverified (harvested, no recorded run). It counts code from repositories linked to that row's paper, not this result: the row's number was not reproduced and nothing here is a correctness claim. The other cell texts mean no graph line for the row: "linked, not harvested" (the archive links code, Syntology has not harvested it), "no code linked" (no code link in the archive), "not matched" (the row's paper URL matched no paper on this site). 23 rows have a graph line, from 20 distinct papers; 18 rows (16 papers) have at least one sample that ran. Counting each paper once: Syntology ran 85 of 239 samples; 154 unverified. Separately, 43 of those 239 are pointer-only (licence): the site points at that code rather than redistributing it, a licence property recorded for ran and unverified samples alike; each cell's tooltip carries the row's own pointer-only count. Read from the graph 2026-09-24. Per-sample status is on the paper page.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections