# Grammatical Error Correction on NLPCC word-level (test) Source: [Grammatical Error Correction on NLPCC word-level (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:58:23.643Z Evaluation of Grammatical Error Correction performance on the test split of the NLPCC word-level dataset, measured using Precision (P), Recall (R), and F0.5 score. A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Grammatical Error Correction | | Dataset (including any stated split) | NLPCC word-level (test) | | Metric | Precision | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Precision](): Higher is better; unit: PERCENT. - [Recall](): Higher is better; unit: PERCENT. - [F0.5 Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Precision | Result date | Report paper | | --- | --- | --- | --- | | Traditional Voting | 64.51 | Not recorded | 68fa6f7f1e209131cea45fce | | Traditional Voting | 58.39 | Not recorded | 68fa6f7f1e209131cea45fce | | seq2seq-1 | 46.17 | Not recorded | 68fa6f7f1e209131cea45fce | | Traditional Voting | 45.58 | Not recorded | 68fa6f7f1e209131cea45fce | | Average of 4 | 43.52 | Not recorded | 68fa6f7f1e209131cea45fce | | seq2seq-2 | 43.4 | Not recorded | 68fa6f7f1e209131cea45fce | | seq2edit-1 | 43.08 | Not recorded | 68fa6f7f1e209131cea45fce | | Sentence-level Ensemble | 42.24 | Not recorded | 68fa6f7f1e209131cea45fce | | Sentence-level Ensemble | 41.94 | Not recorded | 68fa6f7f1e209131cea45fce | | seq2edit-2 | 41.41 | Not recorded | 68fa6f7f1e209131cea45fce | | Sentence-level Ensemble | 41.38 | Not recorded | 68fa6f7f1e209131cea45fce | | Edit-combination Ensemble | 40.5 | Not recorded | 68fa6f7f1e209131cea45fce | | Edit-combination Ensemble | 40.11 | Not recorded | 68fa6f7f1e209131cea45fce | | Edit-level Ensemble | 40.07 | Not recorded | 68fa6f7f1e209131cea45fce | | Edit-level Ensemble | 39.44 | Not recorded | 68fa6f7f1e209131cea45fce | | Edit-combination Ensemble | 37.56 | Not recorded | 68fa6f7f1e209131cea45fce | | Edit-level Ensemble | 36.69 | Not recorded | 68fa6f7f1e209131cea45fce | ### Traditional Voting Precision: 64.51. Result ID: 6990c60d3a6c314849032c00. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Threshold T":4}. This record does not assert independent reproduction. ### Traditional Voting Precision: 58.39. Result ID: 6990c60d3a6c314849032bff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Threshold T":3}. This record does not assert independent reproduction. ### seq2seq-1 Precision: 46.17. Result ID: 6990c60d3a6c314849032bf9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Strategy":"Single Model"}. This record does not assert independent reproduction. ### Traditional Voting Precision: 45.58. Result ID: 6990c60d3a6c314849032bfe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Threshold T":2}. This record does not assert independent reproduction. ### Average of 4 Precision: 43.52. Result ID: 6990c60d3a6c314849032bfd. [Reporting paper]() Recorded evaluation configuration: {"Strategy":"Single Model Ensemble"}. This record does not assert independent reproduction. ### seq2seq-2 Precision: 43.4. Result ID: 6990c60d3a6c314849032bfa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Strategy":"Single Model"}. This record does not assert independent reproduction. ### seq2edit-1 Precision: 43.08. Result ID: 6990c60d3a6c314849032bfb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Strategy":"Single Model"}. This record does not assert independent reproduction. ### Sentence-level Ensemble Precision: 42.24. Result ID: 6990c60d3a6c314849032c02. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"MacBERT-base-Chinese"}. This record does not assert independent reproduction. ### Sentence-level Ensemble Precision: 41.94. Result ID: 6990c60d3a6c314849032c03. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"GPT2-Chinese"}. This record does not assert independent reproduction. ### seq2edit-2 Precision: 41.41. Result ID: 6990c60d3a6c314849032bfc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Strategy":"Single Model"}. This record does not assert independent reproduction. ### Sentence-level Ensemble Precision: 41.38. Result ID: 6990c60d3a6c314849032c01. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"BERT-base-Chinese"}. This record does not assert independent reproduction. ### Edit-combination Ensemble Precision: 40.5. Result ID: 6990c60d3a6c314849032c09. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"GPT2-Chinese"}. This record does not assert independent reproduction. ### Edit-combination Ensemble Precision: 40.11. Result ID: 6990c60d3a6c314849032c08. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"MacBERT-base-Chinese"}. This record does not assert independent reproduction. ### Edit-level Ensemble Precision: 40.07. Result ID: 6990c60d3a6c314849032c05. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"MacBERT-base-Chinese"}. This record does not assert independent reproduction. ### Edit-level Ensemble Precision: 39.44. Result ID: 6990c60d3a6c314849032c06. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"GPT2-Chinese"}. This record does not assert independent reproduction. ### Edit-combination Ensemble Precision: 37.56. Result ID: 6990c60d3a6c314849032c07. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"BERT-base-Chinese"}. This record does not assert independent reproduction. ### Edit-level Ensemble Precision: 36.69. Result ID: 6990c60d3a6c314849032c04. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"PLM Backbone":"BERT-base-Chinese"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.