{"id":7047,"identifier":"DVN/HNA7QO","persistentUrl":"https://doi.org/10.34934/DVN/HNA7QO","protocol":"doi","authority":"10.34934","publisher":"Social Sciences and Digital Humanities Archive – SODHA","publicationDate":"2026-08-25","storageIdentifier":"file://10.34934/DVN/HNA7QO","datasetVersion":{"id":758,"datasetId":7047,"datasetPersistentId":"doi:10.34934/DVN/HNA7QO","storageIdentifier":"file://10.34934/DVN/HNA7QO","versionNumber":1,"versionMinorNumber":0,"versionState":"RELEASED","UNF":"UNF:6:MemhfA5Fl4s1tvLWf4VR2A==","lastUpdateTime":"2026-08-25T14:08:54Z","releaseTime":"2026-08-25T14:08:54Z","createTime":"2026-06-22T17:46:29Z","termsOfAccess":"These files contain training and evaluation data derived from nineteenth-century letters from the Navez Project. To protect access conditions associated with the source material, the files are restricted. Researchers may request access for non-commercial scholarly use. Requests will be evaluated on a case-by-case basis.","fileAccessRequest":true,"metadataBlocks":{"citation":{"displayName":"Citation Metadata","name":"citation","fields":[{"typeName":"title","multiple":false,"typeClass":"primitive","value":"Replication Data for: From nineteenth-century letters to entities: \"a NER pipeline for French correspondence and its methodological lessons\" - Article for Digital Humanities Benelux Journal"},{"typeName":"author","multiple":true,"typeClass":"compound","value":[{"authorName":{"typeName":"authorName","multiple":false,"typeClass":"primitive","value":"Zuzana Černáková"},"authorAffiliation":{"typeName":"authorAffiliation","multiple":false,"typeClass":"primitive","value":"Department of History, Ghent University; Digital Research Lab, KBR"}},{"authorName":{"typeName":"authorName","multiple":false,"typeClass":"primitive","value":"Fien Messens"},"authorAffiliation":{"typeName":"authorAffiliation","multiple":false,"typeClass":"primitive","value":"Department of History, Ghent University; KBR"}},{"authorName":{"typeName":"authorName","multiple":false,"typeClass":"primitive","value":"Tess Dejaeghere"},"authorAffiliation":{"typeName":"authorAffiliation","multiple":false,"typeClass":"primitive","value":"Ghent Centre for Digital Humanities; Language Translation and Technology Team, Ghent University"}},{"authorName":{"typeName":"authorName","multiple":false,"typeClass":"primitive","value":"Julie M. Birkholz"},"authorAffiliation":{"typeName":"authorAffiliation","multiple":false,"typeClass":"primitive","value":"Ghent Centre for Digital Humanities; Department of History, Ghent University; Digital Research Lab, KBR"}}]},{"typeName":"datasetContact","multiple":true,"typeClass":"compound","value":[{"datasetContactName":{"typeName":"datasetContactName","multiple":false,"typeClass":"primitive","value":"Messens, Fien"},"datasetContactAffiliation":{"typeName":"datasetContactAffiliation","multiple":false,"typeClass":"primitive","value":"KBR"}}]},{"typeName":"dsDescription","multiple":true,"typeClass":"compound","value":[{"dsDescriptionValue":{"typeName":"dsDescriptionValue","multiple":false,"typeClass":"primitive","value":"This dataset accompanies the article From nineteenth-century letters to entities: A Named Entity Recognition pipeline for French correspondence and its methodological lessons. It contains the input data, preprocessing scripts, analysis notebooks, evaluation outputs, and experimental results used to develop and evaluate Named Entity Recognition (NER) workflows for nineteenth-century French correspondence from the François-Joseph Navez corpus (KBR – Royal Library of Belgium). The dataset documents the complete experimental workflow, from the construction of a manually annotated gold-standard corpus to the evaluation of off-the-shelf spaCy models, a custom-trained spaCy model, and transformer-based models (CamemBERT, CamemBERTav2, D'AlemBERT and Europeana BERT). The repository is organised into four directories. Input contains the Label Studio annotation exports, the preprocessed gold-standard corpus, and the training, development and test splits used throughout the experiments. Notebooks provides the Jupyter notebooks implementing the preprocessing, training and evaluation workflows. Output contains the evaluation results of the spaCy experiments, including model predictions, overall and per-entity performance metrics, mismatch analyses and summary tables. Results_transformers contains the outputs of the transformer-based experiments, including hyperparameter optimisation, cross-validation results, learning curves, statistical significance tests, prediction files and qualitative error analyses. Together, these files provide the complete computational workflow and all intermediate and final outputs required to reproduce the analyses presented in the accompanying publication."}}]},{"typeName":"subject","multiple":true,"typeClass":"controlledVocabulary","value":["Arts and Humanities"]},{"typeName":"keyword","multiple":true,"typeClass":"compound","value":[{"keywordValue":{"typeName":"keywordValue","multiple":false,"typeClass":"primitive","value":"Named Entity Recognition (NER)"}},{"keywordValue":{"typeName":"keywordValue","multiple":false,"typeClass":"primitive","value":"Natural Language Processing (NLP)"}},{"keywordValue":{"typeName":"keywordValue","multiple":false,"typeClass":"primitive","value":"Historical NLP"}}]},{"typeName":"topicClassification","multiple":true,"typeClass":"compound","value":[{"topicClassVocab":{"typeName":"topicClassVocab","multiple":false,"typeClass":"primitive","value":"CESSDA Topic Classification"},"topicClassVocabURI":{"typeName":"topicClassVocabURI","multiple":false,"typeClass":"primitive","value":"https://vocabularies.cessda.eu/vocabulary/TopicClassification"}}]},{"typeName":"language","multiple":true,"typeClass":"controlledVocabulary","value":["English"]},{"typeName":"depositor","multiple":false,"typeClass":"primitive","value":"Messens, Fien"},{"typeName":"dateOfDeposit","multiple":false,"typeClass":"primitive","value":"2026-06-22"}]}},"files":[{"description":"Preprocessing notebook that converts Label Studio JSON exports into the structured gold-standard corpus used throughout the experiments. The workflow merges annotation projects, extracts entity annotations, aligns character offsets, generates IOB labels, removes unsupported nested entities, and creates the training, development and test datasets used for the NER experiments.","label":"00_navez_gold_standard_preprocessing_json2df.ipynb","restricted":false,"directoryLabel":"Notebooks","version":2,"datasetVersionId":758,"dataFile":{"id":7083,"persistentId":"","pidURL":"","filename":"00_navez_gold_standard_preprocessing_json2df.ipynb","contentType":"application/x-ipynb+json","filesize":476248,"description":"Preprocessing notebook that converts Label Studio JSON exports into the structured gold-standard corpus used throughout the experiments. The workflow merges annotation projects, extracts entity annotations, aligns character offsets, generates IOB labels, removes unsupported nested entities, and creates the training, development and test datasets used for the NER experiments.","storageIdentifier":"file://19ef0aec6e6-3402a7beb950","rootDataFileId":-1,"md5":"6062b90f470e79ad4062aa678de087d3","checksum":{"type":"MD5","value":"6062b90f470e79ad4062aa678de087d3"},"creationDate":"2026-06-22"}},{"description":"Implements Workflow 1 described in the article by evaluating the off-the-shelf spaCy fr_core_news_sm model using only the isolated Named Entity Recognition (NER) component. The notebook evaluates the model on the held-out test set without additional domain-specific training.","label":"01a_navez_NER_spacy_off_the_shelf_no_dk_ner_pipe_sm.ipynb","restricted":false,"directoryLabel":"Notebooks","version":2,"datasetVersionId":758,"dataFile":{"id":7074,"persistentId":"","pidURL":"","filename":"01a_navez_NER_spacy_off_the_shelf_no_dk_ner_pipe_sm.ipynb","contentType":"application/x-ipynb+json","filesize":205854,"description":"Implements Workflow 1 described in the article by evaluating the off-the-shelf spaCy fr_core_news_sm model using only the isolated Named Entity Recognition (NER) component. The notebook evaluates the model on the held-out test set without additional domain-specific training.","storageIdentifier":"file://19ef0aec6f4-26dc7e193221","rootDataFileId":-1,"md5":"fafce906de349b467636197fa64a4d95","checksum":{"type":"MD5","value":"fafce906de349b467636197fa64a4d95"},"creationDate":"2026-06-22"}},{"description":"Implements Workflow 1 using the larger fr_core_news_lg model. The notebook evaluates the isolated NER component on the historical correspondence corpus and compares its performance with the smaller spaCy model.","label":"01b_navez_NER_spacy_off_the_shelf_no_dk_ner_pipe_lg.ipynb","restricted":false,"directoryLabel":"Notebooks","version":2,"datasetVersionId":758,"dataFile":{"id":7105,"persistentId":"","pidURL":"","filename":"01b_navez_NER_spacy_off_the_shelf_no_dk_ner_pipe_lg.ipynb","contentType":"application/x-ipynb+json","filesize":254302,"description":"Implements Workflow 1 using the larger fr_core_news_lg model. The notebook evaluates the isolated NER component on the historical correspondence corpus and compares its performance with the smaller spaCy model.","storageIdentifier":"file://19ef0aec700-9be01b61b903","rootDataFileId":-1,"md5":"4d33ecdf3831757bac0319676a57bb2a","checksum":{"type":"MD5","value":"4d33ecdf3831757bac0319676a57bb2a"},"creationDate":"2026-06-22"}},{"description":"Implements Workflow 2 by using the complete spaCy fr_core_news_sm pipeline, including all NLP components. The notebook assesses whether embedding the NER component within the full pipeline influences recognition performance.","label":"02a_navez_NER_spacy_off_the_shelf_no_dk_nlp_pipe_sm.ipynb","restricted":false,"directoryLabel":"Notebooks","version":4,"datasetVersionId":758,"dataFile":{"id":7100,"persistentId":"","pidURL":"","filename":"02a_navez_NER_spacy_off_the_shelf_no_dk_nlp_pipe_sm.ipynb","contentType":"application/x-ipynb+json","filesize":204787,"description":"Implements Workflow 2 by using the complete spaCy fr_core_news_sm pipeline, including all NLP components. The notebook assesses whether embedding the NER component within the full pipeline influences recognition performance.","storageIdentifier":"file://19ef0aec70d-86cadb948538","rootDataFileId":-1,"md5":"2998c19d3448da4d9ed7a7776de6e9dc","checksum":{"type":"MD5","value":"2998c19d3448da4d9ed7a7776de6e9dc"},"creationDate":"2026-06-22"}},{"description":"Implements Workflow 2 using the complete fr_core_news_lg pipeline. Results are compared with Workflow 1 to evaluate the effect of the full NLP pipeline on NER performance.","label":"02b_navez_NER_spacy_off_the_shelf_no_dk_nlp_pipe_lg.ipynb","restricted":false,"directoryLabel":"Notebooks","version":2,"datasetVersionId":758,"dataFile":{"id":7092,"persistentId":"","pidURL":"","filename":"02b_navez_NER_spacy_off_the_shelf_no_dk_nlp_pipe_lg.ipynb","contentType":"application/x-ipynb+json","filesize":261866,"description":"Implements Workflow 2 using the complete fr_core_news_lg pipeline. Results are compared with Workflow 1 to evaluate the effect of the full NLP pipeline on NER performance.","storageIdentifier":"file://19ef0aec71c-9ccc23ca5b11","rootDataFileId":-1,"md5":"1e35b33c8202285ec93bb0d8508ec8ff","checksum":{"type":"MD5","value":"1e35b33c8202285ec93bb0d8508ec8ff"},"creationDate":"2026-06-22"}},{"description":"Implements Workflow 3 described in the article. The notebook fine-tunes fr_core_news_lg on the Navez gold-standard corpus by extending the default spaCy model with the domain-specific entity types ART, EXH, GRP and LETT, in addition to the standard PER, LOC and ORG categories. It trains and evaluates the custom NER model on the historical correspondence corpus.","label":"03_navez_NER_spacy_custom-trained_dk.ipynb","restricted":false,"directoryLabel":"Notebooks","version":2,"datasetVersionId":758,"dataFile":{"id":7099,"persistentId":"","pidURL":"","filename":"03_navez_NER_spacy_custom-trained_dk.ipynb","contentType":"application/x-ipynb+json","filesize":589345,"description":"Implements Workflow 3 described in the article. The notebook fine-tunes fr_core_news_lg on the Navez gold-standard corpus by extending the default spaCy model with the domain-specific entity types ART, EXH, GRP and LETT, in addition to the standard PER, LOC and ORG categories. It trains and evaluates the custom NER model on the historical correspondence corpus.","storageIdentifier":"file://19ef0aec72b-e7a668c62789","rootDataFileId":-1,"md5":"5b3ef0c547261ba0b3323ce0fd2ca50b","checksum":{"type":"MD5","value":"5b3ef0c547261ba0b3323ce0fd2ca50b"},"creationDate":"2026-06-22"}},{"description":"JSON file containing the final hyperparameter configuration selected for each transformer model (CamemBERT, CamemBERTav2, D'AlemBERT and Europeana BERT) after hyperparameter optimisation.","label":"best_hps.json","restricted":false,"directoryLabel":"Results_transformers","version":2,"datasetVersionId":758,"dataFile":{"id":7050,"persistentId":"","pidURL":"","filename":"best_hps.json","contentType":"application/json","filesize":474,"description":"JSON file containing the final hyperparameter configuration selected for each transformer model (CamemBERT, CamemBERTav2, D'AlemBERT and Europeana BERT) after hyperparameter optimisation.","storageIdentifier":"file://19ef0aec915-8fc92299a705","rootDataFileId":-1,"md5":"a34a656006d26bd49168cb552f29955f","checksum":{"type":"MD5","value":"a34a656006d26bd49168cb552f29955f"},"creationDate":"2026-06-22"}},{"description":"Consolidated spreadsheet containing the evaluation scores of all experiments performed in this study. Facilitates comparison between off-the-shelf spaCy models, the custom-trained spaCy model, and the transformer-based models across different evaluation scenarios and entity types.","label":"combined_eval_scores3.tab","restricted":false,"directoryLabel":"Output","version":4,"datasetVersionId":758,"dataFile":{"id":7093,"persistentId":"","pidURL":"","filename":"combined_eval_scores3.tab","contentType":"text/tab-separated-values","filesize":1466,"description":"Consolidated spreadsheet containing the evaluation scores of all experiments performed in this study. Facilitates comparison between off-the-shelf spaCy models, the custom-trained spaCy model, and the transformer-based models across different evaluation scenarios and entity types.","storageIdentifier":"file://19ef09f7d6c-91daf16abd17","originalFileFormat":"application/vnd.openxmlformats-officedocument.spreadsheetml.sheet","originalFormatLabel":"MS Excel Spreadsheet","originalFileSize":37013,"originalFileName":"combined_eval_scores3.xlsx","UNF":"UNF:6:IHdDRnEBZ4B81/KMQBqVKA==","rootDataFileId":-1,"md5":"f5d63bfa7f4494f86d191eeed39abee0","checksum":{"type":"MD5","value":"f5d63bfa7f4494f86d191eeed39abee0"},"creationDate":"2026-06-22"}},{"description":"Cross-validation results for the transformer-based NER models. Reports evaluation metrics across five training runs with the best hyperparameters per model to assess model robustness.","label":"cv_results.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7103,"persistentId":"","pidURL":"","filename":"cv_results.tab","contentType":"text/tab-separated-values","filesize":1817,"description":"Cross-validation results for the transformer-based NER models. Reports evaluation metrics across five training runs with the best hyperparameters per model to assess model robustness.","storageIdentifier":"file://19ef0aec942-2206cc7524ad","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":1826,"originalFileName":"cv_results.csv","UNF":"UNF:6:72/ohGHOKcR5yXwDUXrubA==","rootDataFileId":-1,"md5":"528d9f3b64ef9f8fb0ab441dc14c3389","checksum":{"type":"MD5","value":"528d9f3b64ef9f8fb0ab441dc14c3389"},"creationDate":"2026-06-22"}},{"description":"Development (validation) partition of the corpus used for hyperparameter optimisation, model selection and intermediate evaluation during NER model training.","label":"development_set-2.csv","restricted":true,"directoryLabel":"Input","version":3,"datasetVersionId":758,"dataFile":{"id":7091,"persistentId":"","pidURL":"","filename":"development_set-2.csv","contentType":"text/csv","filesize":171449,"description":"Development (validation) partition of the corpus used for hyperparameter optimisation, model selection and intermediate evaluation during NER model training.","storageIdentifier":"file://19ef097e67d-3d647c3e29b1","rootDataFileId":-1,"md5":"e2f0fbe2ba2975b0de463eb734e611ab","checksum":{"type":"MD5","value":"e2f0fbe2ba2975b0de463eb734e611ab"},"creationDate":"2026-06-22"}},{"description":"Python Pickle version of the development dataset used in the training workflows.","label":"development_set-2.pkl","restricted":true,"directoryLabel":"Input","version":2,"datasetVersionId":758,"dataFile":{"id":7085,"persistentId":"","pidURL":"","filename":"development_set-2.pkl","contentType":"application/octet-stream","filesize":8619861,"description":"Python Pickle version of the development dataset used in the training workflows.","storageIdentifier":"file://19ef097edca-d196fa1610f5","rootDataFileId":-1,"md5":"8d104a218fd93d5b17d6bf1671104d10","checksum":{"type":"MD5","value":"8d104a218fd93d5b17d6bf1671104d10"},"creationDate":"2026-06-22"}},{"description":"Prediction errors produced by the fine-tuned CamemBERTav2 model. Used for qualitative comparison with the other transformer-based NER models.","label":"errors_camembertav2.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7080,"persistentId":"","pidURL":"","filename":"errors_camembertav2.tab","contentType":"text/tab-separated-values","filesize":14153,"description":"Prediction errors produced by the fine-tuned CamemBERTav2 model. Used for qualitative comparison with the other transformer-based NER models.","storageIdentifier":"file://19ef0aec939-ec3185ef6acc","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":12350,"originalFileName":"errors_camembertav2.csv","UNF":"UNF:6:dX1zhbyrPPk1DreziVEOHw==","rootDataFileId":-1,"md5":"c57cc214817150673d072dda1a83eaf8","checksum":{"type":"MD5","value":"c57cc214817150673d072dda1a83eaf8"},"creationDate":"2026-06-22"}},{"description":"Prediction errors produced by the fine-tuned CamemBERT model. Documents incorrectly recognised, partially recognised, missed and spurious entities for qualitative error analysis.","label":"errors_camembert.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7055,"persistentId":"","pidURL":"","filename":"errors_camembert.tab","contentType":"text/tab-separated-values","filesize":44388,"description":"Prediction errors produced by the fine-tuned CamemBERT model. Documents incorrectly recognised, partially recognised, missed and spurious entities for qualitative error analysis.","storageIdentifier":"file://19ef0aec8f6-f077ac5baa1d","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":38166,"originalFileName":"errors_camembert.csv","UNF":"UNF:6:LKJL6zjeIoC7+p0mefydwg==","rootDataFileId":-1,"md5":"80c88da3c27bf7aa7aa794a7faadd5ac","checksum":{"type":"MD5","value":"80c88da3c27bf7aa7aa794a7faadd5ac"},"creationDate":"2026-06-22"}},{"description":"Prediction errors produced by the fine-tuned D'AlemBERT model. Documents representative recognition errors analysed in the accompanying study.","label":"errors_dalembert.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7106,"persistentId":"","pidURL":"","filename":"errors_dalembert.tab","contentType":"text/tab-separated-values","filesize":16404,"description":"Prediction errors produced by the fine-tuned D'AlemBERT model. Documents representative recognition errors analysed in the accompanying study.","storageIdentifier":"file://19ef0aec91c-5e1fc737bc36","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":14284,"originalFileName":"errors_dalembert.csv","UNF":"UNF:6:+TqzKZgt8g2h9mleixtmLw==","rootDataFileId":-1,"md5":"8e73a5ecd7cd398789830384c3c18cba","checksum":{"type":"MD5","value":"8e73a5ecd7cd398789830384c3c18cba"},"creationDate":"2026-06-22"}},{"description":"Prediction errors produced by the fine-tuned Europeana BERT model. Supports the qualitative evaluation and comparison of transformer-based Named Entity Recognition models.","label":"errors_europeana.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7048,"persistentId":"","pidURL":"","filename":"errors_europeana.tab","contentType":"text/tab-separated-values","filesize":14287,"description":"Prediction errors produced by the fine-tuned Europeana BERT model. Supports the qualitative evaluation and comparison of transformer-based Named Entity Recognition models.","storageIdentifier":"file://19ef0aec8e2-e7b668799c71","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":12454,"originalFileName":"errors_europeana.csv","UNF":"UNF:6:r0qXPIXovg15bXXLwViguQ==","rootDataFileId":-1,"md5":"70955fa4a8d92be4d003320c069a785d","checksum":{"type":"MD5","value":"70955fa4a8d92be4d003320c069a785d"},"creationDate":"2026-06-22"}},{"description":"Table of the final evaluation results obtained on the held-out test set for all transformer-based models under different Nervaluate scenarios. Provides the principal performance metrics reported in the accompanying article and supporting the quantitative comparison of transformer-based models.","label":"headline_counter_table.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7077,"persistentId":"","pidURL":"","filename":"headline_counter_table.tab","contentType":"text/tab-separated-values","filesize":1050,"description":"Table of the final evaluation results obtained on the held-out test set for all transformer-based models under different Nervaluate scenarios. Provides the principal performance metrics reported in the accompanying article and supporting the quantitative comparison of transformer-based models.","storageIdentifier":"file://19ef0aec969-f7fcbb56461d","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":1077,"originalFileName":"headline_counter_table.csv","UNF":"UNF:6:2raSn3MgN38lZjzDY60qZA==","rootDataFileId":-1,"md5":"578544a6c5b336fac13ee9940b6e2c0d","checksum":{"type":"MD5","value":"578544a6c5b336fac13ee9940b6e2c0d"},"creationDate":"2026-06-22"}},{"description":"Summary table of the cross-validation results for all transformer models, facilitating comparison of their overall performance during model selection.","label":"headline_cv_summary.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7056,"persistentId":"","pidURL":"","filename":"headline_cv_summary.tab","contentType":"text/tab-separated-values","filesize":206,"description":"Summary table of the cross-validation results for all transformer models, facilitating comparison of their overall performance during model selection.","storageIdentifier":"file://19ef0aec95c-6a2076fa6e61","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":219,"originalFileName":"headline_cv_summary.csv","UNF":"UNF:6:LwsPP4NHhKrvsotyYYTR+w==","rootDataFileId":-1,"md5":"c4646c8d3455a21e67d7f3e1ab630a1c","checksum":{"type":"MD5","value":"c4646c8d3455a21e67d7f3e1ab630a1c"},"creationDate":"2026-06-22"}},{"description":"Summary of the final evaluation results obtained on the held-out test set for all transformer-based models averaged across three different seeds. ","label":"headline_test_summary.csv","restricted":false,"directoryLabel":"Results_transformers","version":3,"datasetVersionId":758,"dataFile":{"id":7090,"persistentId":"","pidURL":"","filename":"headline_test_summary.csv","contentType":"text/csv","filesize":291,"description":"Summary of the final evaluation results obtained on the held-out test set for all transformer-based models averaged across three different seeds. ","storageIdentifier":"file://19ef0aec901-35e22ffe0d74","rootDataFileId":-1,"md5":"a6b7319263f4f7606f6483082613f61f","checksum":{"type":"MD5","value":"a6b7319263f4f7606f6483082613f61f"},"creationDate":"2026-06-22"}},{"description":"Results of the hyperparameter optimisation experiments performed for the transformer-based Named Entity Recognition models. Records the tested hyperparameter combinations and their corresponding evaluation scores used to identify the optimal training configuration.","label":"hp_search.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7107,"persistentId":"","pidURL":"","filename":"hp_search.tab","contentType":"text/tab-separated-values","filesize":4467,"description":"Results of the hyperparameter optimisation experiments performed for the transformer-based Named Entity Recognition models. Records the tested hyperparameter combinations and their corresponding evaluation scores used to identify the optimal training configuration.","storageIdentifier":"file://19ef0aec94b-9a7948d36b29","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":4371,"originalFileName":"hp_search.csv","UNF":"UNF:6:zu0B79FNEke0HpLSL1WIKw==","rootDataFileId":-1,"md5":"82d61967030768ddb8e823de7b37cb48","checksum":{"type":"MD5","value":"82d61967030768ddb8e823de7b37cb48"},"creationDate":"2026-06-22"}},{"description":"Performance metrics recorded during model training across successive epochs. Used to analyse convergence behaviour, model learning dynamics and training stability.","label":"learning_curve.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7098,"persistentId":"","pidURL":"","filename":"learning_curve.tab","contentType":"text/tab-separated-values","filesize":186,"description":"Performance metrics recorded during model training across successive epochs. Used to analyse convergence behaviour, model learning dynamics and training stability.","storageIdentifier":"file://19ef0aec8c9-b337bdc8b8da","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":234,"originalFileName":"learning_curve.csv","UNF":"UNF:6:fXKPhq8vFAp9d+v7Jlmpsg==","rootDataFileId":-1,"md5":"8edf4d3180870f2d1d95e51223856800","checksum":{"type":"MD5","value":"8edf4d3180870f2d1d95e51223856800"},"creationDate":"2026-06-22"}},{"description":"Evaluation results stratified by document length. Used to assess the influence of letter length on Named Entity Recognition performance across the transformer-based models.","label":"length_stratified.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7073,"persistentId":"","pidURL":"","filename":"length_stratified.tab","contentType":"text/tab-separated-values","filesize":994,"description":"Evaluation results stratified by document length. Used to assess the influence of letter length on Named Entity Recognition performance across the transformer-based models.","storageIdentifier":"file://19ef0aec8ed-b944f4c76e6a","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":976,"originalFileName":"length_stratified.csv","UNF":"UNF:6:/TrpFwgbWWGaD2ZI1dFDKQ==","rootDataFileId":-1,"md5":"291f92dcd2debe65b8d109b66a3cb10b","checksum":{"type":"MD5","value":"291f92dcd2debe65b8d109b66a3cb10b"},"creationDate":"2026-06-22"}},{"description":"Summary statistics describing the distribution of document subword lengths in the  evaluation corpus splitsand their relationship to model performance. Used to assess the number of letters affected by preprocessing. ","label":"length_summary.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7064,"persistentId":"","pidURL":"","filename":"length_summary.tab","contentType":"text/tab-separated-values","filesize":533,"description":"Summary statistics describing the distribution of document subword lengths in the  evaluation corpus splitsand their relationship to model performance. Used to assess the number of letters affected by preprocessing. ","storageIdentifier":"file://19ef0aec973-b08cbe3d619f","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":549,"originalFileName":"length_summary.csv","UNF":"UNF:6:XKylVdddSgIVxmbYCdrnFg==","rootDataFileId":-1,"md5":"469bd6aa506a2f3cc0c1a93968eae40a","checksum":{"type":"MD5","value":"469bd6aa506a2f3cc0c1a93968eae40a"},"creationDate":"2026-06-22"}},{"description":"Preprocessed version of the manually annotated corpus after cleaning and conversion from Label Studio. Contains token- and entity-level information used for model training and evaluation. ","label":"navez_gs_preprocessed-2.csv","restricted":true,"directoryLabel":"Input","version":2,"datasetVersionId":758,"dataFile":{"id":7079,"persistentId":"","pidURL":"","filename":"navez_gs_preprocessed-2.csv","contentType":"text/csv","filesize":1039636,"description":"Preprocessed version of the manually annotated corpus after cleaning and conversion from Label Studio. Contains token- and entity-level information used for model training and evaluation. ","storageIdentifier":"file://19ef097ef49-34c96a3b58a6","rootDataFileId":-1,"md5":"0469ba07b6aa2663770de059d36ef03f","checksum":{"type":"MD5","value":"0469ba07b6aa2663770de059d36ef03f"},"creationDate":"2026-06-22"}},{"description":"Python pickle version of the preprocessed gold-standard corpus, preserving data structures used directly in the notebooks.","label":"navez_gs_preprocessed-2.pkl","restricted":true,"directoryLabel":"Input","version":2,"datasetVersionId":758,"dataFile":{"id":7104,"persistentId":"","pidURL":"","filename":"navez_gs_preprocessed-2.pkl","contentType":"application/octet-stream","filesize":34654596,"description":"Python pickle version of the preprocessed gold-standard corpus, preserving data structures used directly in the notebooks.","storageIdentifier":"file://19ef0980c0d-7fe4d8a123d6","rootDataFileId":-1,"md5":"3a11affcc28d1dc3ac86f185c3f089a2","checksum":{"type":"MD5","value":"3a11affcc28d1dc3ac86f185c3f089a2"},"creationDate":"2026-06-22"}},{"description":"Notebook for fine-tuning and evaluating French transformer-based models for Named Entity Recognition on the Navez correspondence. It implements the experiments with CamemBERT, CamemBERTav2, D'AlemBERT and Europeana BERT, including weighted cross-entropy training, hyperparameter optimisation and evaluation using the Nervaluate framework.","label":"Navez_ner_finetuning_weightedcrossentropy.ipynb","restricted":false,"directoryLabel":"Notebooks","version":2,"datasetVersionId":758,"dataFile":{"id":7063,"persistentId":"","pidURL":"","filename":"Navez_ner_finetuning_weightedcrossentropy.ipynb","contentType":"application/x-ipynb+json","filesize":4869761,"description":"Notebook for fine-tuning and evaluating French transformer-based models for Named Entity Recognition on the Navez correspondence. It implements the experiments with CamemBERT, CamemBERTav2, D'AlemBERT and Europeana BERT, including weighted cross-entropy training, hyperparameter optimisation and evaluation using the Nervaluate framework.","storageIdentifier":"file://19ef0aec6bd-d731fdaef09d","rootDataFileId":-1,"md5":"8c2912db3ec1743bd4c24c1103a01c99","checksum":{"type":"MD5","value":"8c2912db3ec1743bd4c24c1103a01c99"},"creationDate":"2026-06-22"}},{"description":"Overall evaluation metrics for the custom-trained spaCy model, including the four project-specific entity types introduced during fine-tuning. Reports precision, recall and F1 under the Nervaluate evaluation scenarios.","label":"navez_spacy_custom_trained_evalall-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7071,"persistentId":"","pidURL":"","filename":"navez_spacy_custom_trained_evalall-2.csv","contentType":"text/csv","filesize":424,"description":"Overall evaluation metrics for the custom-trained spaCy model, including the four project-specific entity types introduced during fine-tuning. Reports precision, recall and F1 under the Nervaluate evaluation scenarios.","storageIdentifier":"file://19ef0982b49-fff35870062f","rootDataFileId":-1,"md5":"2431b25f16ced713f3f94efd19eede8d","checksum":{"type":"MD5","value":"2431b25f16ced713f3f94efd19eede8d"},"creationDate":"2026-06-22"}},{"description":"Overall evaluation of the custom-trained model restricted to the standard spaCy entity types (PER, LOC and ORG). This output enables direct comparison with the off-the-shelf spaCy models presented in the article.","label":"navez_spacy_custom_trained_evalall_no_dk-2.tab","restricted":false,"directoryLabel":"Output","version":4,"datasetVersionId":758,"dataFile":{"id":7058,"persistentId":"","pidURL":"","filename":"navez_spacy_custom_trained_evalall_no_dk-2.tab","contentType":"text/tab-separated-values","filesize":462,"description":"Overall evaluation of the custom-trained model restricted to the standard spaCy entity types (PER, LOC and ORG). This output enables direct comparison with the off-the-shelf spaCy models presented in the article.","storageIdentifier":"file://19ef0982b92-79cdfbd0cd69","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":561,"originalFileName":"navez_spacy_custom_trained_evalall_no_dk-2.csv","UNF":"UNF:6:yq7KQEdVC0k1T86KGkSn0g==","rootDataFileId":-1,"md5":"4428457c61e96b94e212184d398149e8","checksum":{"type":"MD5","value":"4428457c61e96b94e212184d398149e8"},"creationDate":"2026-06-22"}},{"description":"Detailed evaluation metrics for each entity type recognised by the custom-trained model, including both standard and domain-specific categories (PER, LOC, ORG, GRP, ART, EXH and LETT).","label":"navez_spacy_custom_trained_evalpertype-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7086,"persistentId":"","pidURL":"","filename":"navez_spacy_custom_trained_evalpertype-2.csv","contentType":"text/csv","filesize":2036,"description":"Detailed evaluation metrics for each entity type recognised by the custom-trained model, including both standard and domain-specific categories (PER, LOC, ORG, GRP, ART, EXH and LETT).","storageIdentifier":"file://19ef0982c0f-4fe011f89ba8","rootDataFileId":-1,"md5":"d770abff6f7b4ac2791c99f080b2ccc9","checksum":{"type":"MD5","value":"d770abff6f7b4ac2791c99f080b2ccc9"},"creationDate":"2026-06-22"}},{"description":"Token-level comparison between predicted and gold-standard IOB labels for the custom-trained spaCy model. Used for the qualitative error analysis discussed in the article.","label":"navez_spacy_custom_trained_mismatches.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7068,"persistentId":"","pidURL":"","filename":"navez_spacy_custom_trained_mismatches.csv","contentType":"text/csv","filesize":1661419,"description":"Token-level comparison between predicted and gold-standard IOB labels for the custom-trained spaCy model. Used for the qualitative error analysis discussed in the article.","storageIdentifier":"file://19ef09f8038-226a3f090987","rootDataFileId":-1,"md5":"bd7431776ef54f320a6345ee2c39b5d0","checksum":{"type":"MD5","value":"bd7431776ef54f320a6345ee2c39b5d0"},"creationDate":"2026-06-22"}},{"description":"Predictions generated by the custom-trained fr_core_news_lg model after fine-tuning on the Navez gold-standard corpus. Includes all project-specific entity types (PER, LOC, ORG, GRP, ART, EXH and LETT).","label":"navez_spacy_custom_trained_results.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7051,"persistentId":"","pidURL":"","filename":"navez_spacy_custom_trained_results.csv","contentType":"text/csv","filesize":159612,"description":"Predictions generated by the custom-trained fr_core_news_lg model after fine-tuning on the Navez gold-standard corpus. Includes all project-specific entity types (PER, LOC, ORG, GRP, ART, EXH and LETT).","storageIdentifier":"file://19ef09f823e-2e630aeb1e94","rootDataFileId":-1,"md5":"519958789ccf7a67f90669ce21103d90","checksum":{"type":"MD5","value":"519958789ccf7a67f90669ce21103d90"},"creationDate":"2026-06-22"}},{"description":"Overall evaluation results for the off-the-shelf fr_core_news_lg model. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","label":"navez_spacy_off_the_shelf_ner_lg_evalall-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7094,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_lg_evalall-2.csv","contentType":"text/csv","filesize":410,"description":"Overall evaluation results for the off-the-shelf fr_core_news_lg model. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","storageIdentifier":"file://19ef0982c4f-65eeabc22872","rootDataFileId":-1,"md5":"27ce89c83902be5851a3c2a3a4c0cf30","checksum":{"type":"MD5","value":"27ce89c83902be5851a3c2a3a4c0cf30"},"creationDate":"2026-06-22"}},{"description":"Evaluation metrics for each entity type recognised by the off-the-shelf fr_core_news_lg model (PER, LOC and ORG). Enables comparison of model performance across entity categories.","label":"navez_spacy_off_the_shelf_ner_lg_evalpertype-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7069,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_lg_evalpertype-2.csv","contentType":"text/csv","filesize":988,"description":"Evaluation metrics for each entity type recognised by the off-the-shelf fr_core_news_lg model (PER, LOC and ORG). Enables comparison of model performance across entity categories.","storageIdentifier":"file://19ef0982ca0-3e66cad4598d","rootDataFileId":-1,"md5":"95c489a1ea58b92b8d94ef795bc284ce","checksum":{"type":"MD5","value":"95c489a1ea58b92b8d94ef795bc284ce"},"creationDate":"2026-06-22"}},{"description":"Token-level comparison between predicted and gold-standard IOB labels for the off-the-shelf fr_core_news_lg model. Used for qualitative error analysis and identification of recurring recognition errors.","label":"navez_spacy_off_the_shelf_ner_lg_mismatches.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7096,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_lg_mismatches.csv","contentType":"text/csv","filesize":2048453,"description":"Token-level comparison between predicted and gold-standard IOB labels for the off-the-shelf fr_core_news_lg model. Used for qualitative error analysis and identification of recurring recognition errors.","storageIdentifier":"file://19ef09f8577-73c12c8d500a","rootDataFileId":-1,"md5":"19e45bf3f0974a05524450f94b2b93da","checksum":{"type":"MD5","value":"19e45bf3f0974a05524450f94b2b93da"},"creationDate":"2026-06-22"}},{"description":"Model predictions generated by the off-the-shelf fr_core_news_lg model using the isolated NER component (Workflow 1). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","label":"navez_spacy_off_the_shelf_ner_lg_results.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7066,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_lg_results.csv","contentType":"text/csv","filesize":160670,"description":"Model predictions generated by the off-the-shelf fr_core_news_lg model using the isolated NER component (Workflow 1). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","storageIdentifier":"file://19ef09f871d-43e76fc4e18f","rootDataFileId":-1,"md5":"7c28c59d7b17430f781573f98a1a5833","checksum":{"type":"MD5","value":"7c28c59d7b17430f781573f98a1a5833"},"creationDate":"2026-06-22"}},{"description":"Overall evaluation results for the off-the-shelf fr_core_news_sm model. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","label":"navez_spacy_off_the_shelf_ner_sm_evalall-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7061,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_sm_evalall-2.csv","contentType":"text/csv","filesize":439,"description":"Overall evaluation results for the off-the-shelf fr_core_news_sm model. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","storageIdentifier":"file://19ef0982cf1-58bb3930873b","rootDataFileId":-1,"md5":"b1e463b5396ee3ecd1eb9256b33d3ccf","checksum":{"type":"MD5","value":"b1e463b5396ee3ecd1eb9256b33d3ccf"},"creationDate":"2026-06-22"}},{"description":"Evaluation metrics for each entity type recognised by the off-the-shelf fr_core_news_sm model (PER, LOC and ORG). Enables comparison of model performance across entity categories.","label":"navez_spacy_off_the_shelf_ner_sm_evalpertype-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7062,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_sm_evalpertype-2.csv","contentType":"text/csv","filesize":1118,"description":"Evaluation metrics for each entity type recognised by the off-the-shelf fr_core_news_sm model (PER, LOC and ORG). Enables comparison of model performance across entity categories.","storageIdentifier":"file://19ef0982d44-2ebbad9386be","rootDataFileId":-1,"md5":"1c6948aab25b26cdc34d89384e66544a","checksum":{"type":"MD5","value":"1c6948aab25b26cdc34d89384e66544a"},"creationDate":"2026-06-22"}},{"description":"Token-level comparison between predicted and gold-standard IOB labels for the off-the-shelf fr_core_news_sm model. Used for qualitative error analysis and identification of recurring recognition errors.","label":"navez_spacy_off_the_shelf_ner_sm_mismatches.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7108,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_sm_mismatches.csv","contentType":"text/csv","filesize":2291265,"description":"Token-level comparison between predicted and gold-standard IOB labels for the off-the-shelf fr_core_news_sm model. Used for qualitative error analysis and identification of recurring recognition errors.","storageIdentifier":"file://19ef09f6fcf-707c396f8e6e","rootDataFileId":-1,"md5":"fdd354911d784ba25ace9972ad447335","checksum":{"type":"MD5","value":"fdd354911d784ba25ace9972ad447335"},"creationDate":"2026-06-22"}},{"description":"Model predictions generated by the off-the-shelf fr_core_news_sm model using the isolated NER component (Workflow 1). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","label":"navez_spacy_off_the_shelf_ner_sm_results.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7059,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_ner_sm_results.csv","contentType":"text/csv","filesize":161008,"description":"Model predictions generated by the off-the-shelf fr_core_news_sm model using the isolated NER component (Workflow 1). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","storageIdentifier":"file://19ef09f717f-6937e3704b14","rootDataFileId":-1,"md5":"5f684b97b3c570eb8ada7751bdb6a075","checksum":{"type":"MD5","value":"5f684b97b3c570eb8ada7751bdb6a075"},"creationDate":"2026-06-22"}},{"description":"Overall evaluation results for the complete fr_core_news_lg NLP pipeline. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","label":"navez_spacy_off_the_shelf_nlp_lg_evalall-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7067,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_lg_evalall-2.csv","contentType":"text/csv","filesize":434,"description":"Overall evaluation results for the complete fr_core_news_lg NLP pipeline. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","storageIdentifier":"file://19ef0982d90-b4a9e4eb178e","rootDataFileId":-1,"md5":"0b166ca4796478707cf53976e9ed3da1","checksum":{"type":"MD5","value":"0b166ca4796478707cf53976e9ed3da1"},"creationDate":"2026-06-22"}},{"description":"Evaluation metrics for each entity type recognised by the complete fr_core_news_lg pipeline (PER, LOC and ORG). Enables comparison with the isolated NER workflow and the custom-trained model.","label":"navez_spacy_off_the_shelf_nlp_lg_evalpertype-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7057,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_lg_evalpertype-2.csv","contentType":"text/csv","filesize":1042,"description":"Evaluation metrics for each entity type recognised by the complete fr_core_news_lg pipeline (PER, LOC and ORG). Enables comparison with the isolated NER workflow and the custom-trained model.","storageIdentifier":"file://19ef0982dd3-2813a033f911","rootDataFileId":-1,"md5":"02002572ce6487ca896a367b54f4d998","checksum":{"type":"MD5","value":"02002572ce6487ca896a367b54f4d998"},"creationDate":"2026-06-22"}},{"description":"Token-level comparison between predicted and gold-standard IOB labels for the complete fr_core_news_lg NLP pipeline. Used for qualitative error analysis and comparison with the isolated NER workflow.","label":"navez_spacy_off_the_shelf_nlp_lg_mismatches.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7049,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_lg_mismatches.csv","contentType":"text/csv","filesize":2032357,"description":"Token-level comparison between predicted and gold-standard IOB labels for the complete fr_core_news_lg NLP pipeline. Used for qualitative error analysis and comparison with the isolated NER workflow.","storageIdentifier":"file://19ef09f74db-90fd4a0fd704","rootDataFileId":-1,"md5":"66e42633a3ae8e4f7ac2aa2bec9ba813","checksum":{"type":"MD5","value":"66e42633a3ae8e4f7ac2aa2bec9ba813"},"creationDate":"2026-06-22"}},{"description":"Model predictions generated by the off-the-shelf fr_core_news_lg model using the complete spaCy NLP pipeline (Workflow 2). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","label":"navez_spacy_off_the_shelf_nlp_lg_results.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7053,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_lg_results.csv","contentType":"text/csv","filesize":160589,"description":"Model predictions generated by the off-the-shelf fr_core_news_lg model using the complete spaCy NLP pipeline (Workflow 2). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","storageIdentifier":"file://19ef09f76d9-47b55a2b53eb","rootDataFileId":-1,"md5":"0c0b15b22c6286d52351ce6df21146e7","checksum":{"type":"MD5","value":"0c0b15b22c6286d52351ce6df21146e7"},"creationDate":"2026-06-22"}},{"description":"Overall evaluation results for the complete fr_core_news_sm NLP pipeline. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","label":"navez_spacy_off_the_shelf_nlp_sm_evalall-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7065,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_sm_evalall-2.csv","contentType":"text/csv","filesize":438,"description":"Overall evaluation results for the complete fr_core_news_sm NLP pipeline. Reports precision, recall and F1 scores under the Strict, Exact, Partial and Type evaluation scenarios using the Nervaluate framework.","storageIdentifier":"file://19ef0982e0d-6a249553d56e","rootDataFileId":-1,"md5":"60738aaa0921c5c4de7de9e63ecca08a","checksum":{"type":"MD5","value":"60738aaa0921c5c4de7de9e63ecca08a"},"creationDate":"2026-06-22"}},{"description":"Evaluation metrics for each entity type recognised by the complete fr_core_news_sm pipeline (PER, LOC and ORG). Enables comparison with the isolated NER workflow.","label":"navez_spacy_off_the_shelf_nlp_sm_evalpertype-2.csv","restricted":false,"directoryLabel":"Output","version":2,"datasetVersionId":758,"dataFile":{"id":7095,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_sm_evalpertype-2.csv","contentType":"text/csv","filesize":1052,"description":"Evaluation metrics for each entity type recognised by the complete fr_core_news_sm pipeline (PER, LOC and ORG). Enables comparison with the isolated NER workflow.","storageIdentifier":"file://19ef0982e60-046fdb366786","rootDataFileId":-1,"md5":"0063ddb26e8f5f36183bac8e2cffb6d2","checksum":{"type":"MD5","value":"0063ddb26e8f5f36183bac8e2cffb6d2"},"creationDate":"2026-06-22"}},{"description":"Token-level comparison between predicted and gold-standard IOB labels for the complete fr_core_news_sm NLP pipeline. Used for the qualitative error analysis and identification of recurring recognition errors.","label":"navez_spacy_off_the_shelf_nlp_sm_mismatches.csv","restricted":true,"directoryLabel":"Output","version":5,"datasetVersionId":758,"dataFile":{"id":7102,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_sm_mismatches.csv","contentType":"text/csv","filesize":2246034,"description":"Token-level comparison between predicted and gold-standard IOB labels for the complete fr_core_news_sm NLP pipeline. Used for the qualitative error analysis and identification of recurring recognition errors.","storageIdentifier":"file://19ef09f7a26-aee98d85f4df","rootDataFileId":-1,"md5":"25c9405590014091e5d08bf4aebb8301","checksum":{"type":"MD5","value":"25c9405590014091e5d08bf4aebb8301"},"creationDate":"2026-06-22"}},{"description":"Model predictions generated by the off-the-shelf fr_core_news_sm model using the complete spaCy NLP pipeline (Workflow 2). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","label":"navez_spacy_off_the_shelf_nlp_sm_results.csv","restricted":true,"directoryLabel":"Output","version":3,"datasetVersionId":758,"dataFile":{"id":7076,"persistentId":"","pidURL":"","filename":"navez_spacy_off_the_shelf_nlp_sm_results.csv","contentType":"text/csv","filesize":160776,"description":"Model predictions generated by the off-the-shelf fr_core_news_sm model using the complete spaCy NLP pipeline (Workflow 2). Contains predicted entities, spans, labels and corresponding gold-standard annotations for the held-out test set.","storageIdentifier":"file://19ef09f7be2-1c95058d6963","rootDataFileId":-1,"md5":"1128164075ad2b33e1902c44b764d000","checksum":{"type":"MD5","value":"1128164075ad2b33e1902c44b764d000"},"creationDate":"2026-06-22"}},{"description":"Aggregated Nervaluate statistics reporting the numbers of correct, incorrect, partial, missed and spurious entity predictions on the held-out test set under the different evaluation scenarios (Strict, Exact, Partial and Type) for all transformer-based models. ","label":"nervaluate_counters_overall.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7075,"persistentId":"","pidURL":"","filename":"nervaluate_counters_overall.tab","contentType":"text/tab-separated-values","filesize":1811,"description":"Aggregated Nervaluate statistics reporting the numbers of correct, incorrect, partial, missed and spurious entity predictions on the held-out test set under the different evaluation scenarios (Strict, Exact, Partial and Type) for all transformer-based models. ","storageIdentifier":"file://19ef0aec930-bfce3ab59464","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":1784,"originalFileName":"nervaluate_counters_overall.csv","UNF":"UNF:6:/3uciuaa/yLyO27c+p2Cyg==","rootDataFileId":-1,"md5":"855504ea906378d3046317854080edf6","checksum":{"type":"MD5","value":"855504ea906378d3046317854080edf6"},"creationDate":"2026-06-22"}},{"description":"Evaluation results for each transformer-based model and each entity type (PER, LOC, ORG, GRP, ART, EXH and LETT) on the held-out test set under the Nervaluate scenarios. It enables detailed comparison of model performance across annotation categories. ","label":"per_tag_test.tab","restricted":false,"directoryLabel":"Results_transformers","version":5,"datasetVersionId":758,"dataFile":{"id":7082,"persistentId":"","pidURL":"","filename":"per_tag_test.tab","contentType":"text/tab-separated-values","filesize":12846,"description":"Evaluation results for each transformer-based model and each entity type (PER, LOC, ORG, GRP, ART, EXH and LETT) on the held-out test set under the Nervaluate scenarios. It enables detailed comparison of model performance across annotation categories. ","storageIdentifier":"file://19ef0aec907-f023a24abc28","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":11923,"originalFileName":"per_tag_test.csv","UNF":"UNF:6:YzGwFRM6nYl9P3yPhkiCEg==","rootDataFileId":-1,"md5":"95920c915449358501f2a7dae71d4b28","checksum":{"type":"MD5","value":"95920c915449358501f2a7dae71d4b28"},"creationDate":"2026-06-22"}},{"description":"Part 1 - JSON export from Label Studio containing manually annotated nineteenth-century French correspondence from the Navez Project. Includes entity annotations, transcriptions and project metadata. Used as source data for preprocessing and model training.","label":"project-237-at-2025-05-25-08-36-f874dc2a-2.json","restricted":true,"directoryLabel":"Input","version":3,"datasetVersionId":758,"dataFile":{"id":7054,"persistentId":"","pidURL":"","filename":"project-237-at-2025-05-25-08-36-f874dc2a-2.json","contentType":"application/json","filesize":203344,"description":"Part 1 - JSON export from Label Studio containing manually annotated nineteenth-century French correspondence from the Navez Project. Includes entity annotations, transcriptions and project metadata. Used as source data for preprocessing and model training.","storageIdentifier":"file://19ef0980d82-3f563b2720a5","rootDataFileId":-1,"md5":"72da0345995240cc733071556d1a647e","checksum":{"type":"MD5","value":"72da0345995240cc733071556d1a647e"},"creationDate":"2026-06-22"}},{"description":"Part 2 - JSON export from Label Studio containing manually annotated nineteenth-century French correspondence from the Navez Project. Includes entity annotations, transcriptions and project metadata. Used as source data for preprocessing and model training.","label":"project-238-at-2025-05-25-08-34-694e26ec-2.json","restricted":true,"directoryLabel":"Input","version":3,"datasetVersionId":758,"dataFile":{"id":7052,"persistentId":"","pidURL":"","filename":"project-238-at-2025-05-25-08-34-694e26ec-2.json","contentType":"application/json","filesize":588923,"description":"Part 2 - JSON export from Label Studio containing manually annotated nineteenth-century French correspondence from the Navez Project. Includes entity annotations, transcriptions and project metadata. Used as source data for preprocessing and model training.","storageIdentifier":"file://19ef0980e40-31ee08295b06","rootDataFileId":-1,"md5":"526f06d4f9ce02a351f068df4162d62b","checksum":{"type":"MD5","value":"526f06d4f9ce02a351f068df4162d62b"},"creationDate":"2026-06-22"}},{"description":"Results of statistical significance tests comparing the performance of the evaluated Named Entity Recognition models. Used to assess whether observed performance differences are statistically meaningful.","label":"significance_tests.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7089,"persistentId":"","pidURL":"","filename":"significance_tests.tab","contentType":"text/tab-separated-values","filesize":559,"description":"Results of statistical significance tests comparing the performance of the evaluated Named Entity Recognition models. Used to assess whether observed performance differences are statistically meaningful.","storageIdentifier":"file://19ef0aec961-c4a11e73653e","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":574,"originalFileName":"significance_tests.csv","UNF":"UNF:6:9YJ3w4uTtbYmfNKxJTh1UA==","rootDataFileId":-1,"md5":"8bf8513f4c7460b5d3d9e2b520e88b51","checksum":{"type":"MD5","value":"8bf8513f4c7460b5d3d9e2b520e88b51"},"creationDate":"2026-06-22"}},{"description":"Summary of the number of annotated entities per entity type in the training, development and test datasets. Used to document the composition of the gold-standard corpus and the experimental data splits.","label":"split_entity_counts.csv","restricted":false,"directoryLabel":"Results_transformers","version":2,"datasetVersionId":758,"dataFile":{"id":7101,"persistentId":"","pidURL":"","filename":"split_entity_counts.csv","contentType":"text/csv","filesize":162,"description":"Summary of the number of annotated entities per entity type in the training, development and test datasets. Used to document the composition of the gold-standard corpus and the experimental data splits.","storageIdentifier":"file://19ef0aec910-56ec01010b16","rootDataFileId":-1,"md5":"ae69733d2f22cb00b53e031a03f99e3f","checksum":{"type":"MD5","value":"ae69733d2f22cb00b53e031a03f99e3f"},"creationDate":"2026-06-22"}},{"description":"Named Entity Recognition predictions generated by the fine-tuned CamemBERTav2 model on the held-out test set. Includes predicted entity spans, labels and gold-standard annotations for model evaluation.","label":"test_predictions_camembertav2.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7097,"persistentId":"","pidURL":"","filename":"test_predictions_camembertav2.tab","contentType":"text/tab-separated-values","filesize":98313,"description":"Named Entity Recognition predictions generated by the fine-tuned CamemBERTav2 model on the held-out test set. Includes predicted entity spans, labels and gold-standard annotations for model evaluation.","storageIdentifier":"file://19ef0aec8cf-ef84afa376be","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":98300,"originalFileName":"test_predictions_camembertav2.csv","UNF":"UNF:6:6PxO9ELAf/ENQQAY98ekCQ==","rootDataFileId":-1,"md5":"8719c1e4cce65c889733eabc70996504","checksum":{"type":"MD5","value":"8719c1e4cce65c889733eabc70996504"},"creationDate":"2026-06-22"}},{"description":"Named Entity Recognition predictions generated by the fine-tuned CamemBERT model on the held-out test set. Includes predicted entity spans, labels and gold-standard annotations for model evaluation.","label":"test_predictions_camembert.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7078,"persistentId":"","pidURL":"","filename":"test_predictions_camembert.tab","contentType":"text/tab-separated-values","filesize":101354,"description":"Named Entity Recognition predictions generated by the fine-tuned CamemBERT model on the held-out test set. Includes predicted entity spans, labels and gold-standard annotations for model evaluation.","storageIdentifier":"file://19ef0aec926-06fdabee3d74","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":101341,"originalFileName":"test_predictions_camembert.csv","UNF":"UNF:6:wwYLN6EuL27q8HvnSckZ9A==","rootDataFileId":-1,"md5":"6eaabd0956bc79783561a84558eb326c","checksum":{"type":"MD5","value":"6eaabd0956bc79783561a84558eb326c"},"creationDate":"2026-06-22"}},{"description":"Named Entity Recognition predictions generated by the fine-tuned D'AlemBERT model on the held-out test set. Includes predicted entity spans, labels and gold-standard annotations for model evaluation.","label":"test_predictions_dalembert.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7088,"persistentId":"","pidURL":"","filename":"test_predictions_dalembert.tab","contentType":"text/tab-separated-values","filesize":98579,"description":"Named Entity Recognition predictions generated by the fine-tuned D'AlemBERT model on the held-out test set. Includes predicted entity spans, labels and gold-standard annotations for model evaluation.","storageIdentifier":"file://19ef0aec8c0-9edeca7c88d4","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":98566,"originalFileName":"test_predictions_dalembert.csv","UNF":"UNF:6:QOnGSP2nZCOfl3Bdo7Wayw==","rootDataFileId":-1,"md5":"ba7d3dc2feda5828181a7dc6eaac70fa","checksum":{"type":"MD5","value":"ba7d3dc2feda5828181a7dc6eaac70fa"},"creationDate":"2026-06-22"}},{"description":"Named Entity Recognition predictions generated by the fine-tuned Europeana BERT model on the held-out test set. Contains predicted entity spans, labels and corresponding gold-standard annotations used for the quantitative and qualitative evaluation of model performance.","label":"test_predictions_europeana.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7072,"persistentId":"","pidURL":"","filename":"test_predictions_europeana.tab","contentType":"text/tab-separated-values","filesize":98398,"description":"Named Entity Recognition predictions generated by the fine-tuned Europeana BERT model on the held-out test set. Contains predicted entity spans, labels and corresponding gold-standard annotations used for the quantitative and qualitative evaluation of model performance.","storageIdentifier":"file://19ef0aec8d9-49d0801564b8","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":98385,"originalFileName":"test_predictions_europeana.csv","UNF":"UNF:6:47AUFPWKuqAoGujhisXcrQ==","rootDataFileId":-1,"md5":"df5312b00c61230de598cd09c1b5458a","checksum":{"type":"MD5","value":"df5312b00c61230de598cd09c1b5458a"},"creationDate":"2026-06-22"}},{"description":"Evaluation results of the transformer-based Named Entity Recognition models across multiple random seeds. Reports the test-set performance of each training run and was used to assess the robustness and reproducibility of the experimental results by calculating mean performance across seeds.","label":"test_scores_per_seed.tab","restricted":false,"directoryLabel":"Results_transformers","version":4,"datasetVersionId":758,"dataFile":{"id":7084,"persistentId":"","pidURL":"","filename":"test_scores_per_seed.tab","contentType":"text/tab-separated-values","filesize":5065,"description":"Evaluation results of the transformer-based Named Entity Recognition models across multiple random seeds. Reports the test-set performance of each training run and was used to assess the robustness and reproducibility of the experimental results by calculating mean performance across seeds.","storageIdentifier":"file://19ef0aec953-00288031ca9c","originalFileFormat":"text/csv","originalFormatLabel":"Comma Separated Values","originalFileSize":4969,"originalFileName":"test_scores_per_seed.csv","UNF":"UNF:6:BULzL9VsIcllUNnVwyPBzg==","rootDataFileId":-1,"md5":"6004c572ef856db8b3dace8f6bf13a63","checksum":{"type":"MD5","value":"6004c572ef856db8b3dace8f6bf13a63"},"creationDate":"2026-06-22"}},{"description":"Held-out test split used exclusively for final evaluation of the trained NER models. ","label":"test_set-2.csv","restricted":true,"directoryLabel":"Input","version":2,"datasetVersionId":758,"dataFile":{"id":7060,"persistentId":"","pidURL":"","filename":"test_set-2.csv","contentType":"text/csv","filesize":120351,"description":"Held-out test split used exclusively for final evaluation of the trained NER models. ","storageIdentifier":"file://19ef0980e9d-61df8f6b9232","rootDataFileId":-1,"md5":"c079a2cd638f0797e2b2ec4e06348e29","checksum":{"type":"MD5","value":"c079a2cd638f0797e2b2ec4e06348e29"},"creationDate":"2026-06-22"}},{"description":"The file contains manually annotated letter transcriptions and their token-level entity labels. For each letter, it includes the manuscript identifier, full transcription, tokenized text, annotated entities and entity types, character and token spans, remapped annotations, and IOB tags.","label":"test_set-2.pkl","restricted":true,"directoryLabel":"Input","version":2,"datasetVersionId":758,"dataFile":{"id":7070,"persistentId":"","pidURL":"","filename":"test_set-2.pkl","contentType":"application/octet-stream","filesize":7107691,"description":"The file contains manually annotated letter transcriptions and their token-level entity labels. For each letter, it includes the manuscript identifier, full transcription, tokenized text, annotated entities and entity types, character and token spans, remapped annotations, and IOB tags.","storageIdentifier":"file://19ef0981480-53bb755c05b6","rootDataFileId":-1,"md5":"ebeb69df35b8eecb39e9ce2342040c8d","checksum":{"type":"MD5","value":"ebeb69df35b8eecb39e9ce2342040c8d"},"creationDate":"2026-06-22"}},{"description":"Held-out test partition of the gold-standard corpus used exclusively for final evaluation of the NER models after training.","label":"training_set-2.csv","restricted":true,"directoryLabel":"Input","version":3,"datasetVersionId":758,"dataFile":{"id":7081,"persistentId":"","pidURL":"","filename":"training_set-2.csv","contentType":"text/csv","filesize":748017,"description":"Held-out test partition of the gold-standard corpus used exclusively for final evaluation of the NER models after training.","storageIdentifier":"file://19ef0981595-61b6f0e16796","rootDataFileId":-1,"md5":"8da5d01228c4818dddce6a97272b18be","checksum":{"type":"MD5","value":"8da5d01228c4818dddce6a97272b18be"},"creationDate":"2026-06-22"}},{"description":"Python Pickle version of the held-out test dataset used in the evaluation pipeline.","label":"training_set-2.pkl","restricted":true,"directoryLabel":"Input","version":3,"datasetVersionId":758,"dataFile":{"id":7087,"persistentId":"","pidURL":"","filename":"training_set-2.pkl","contentType":"application/octet-stream","filesize":26006134,"description":"Python Pickle version of the held-out test dataset used in the evaluation pipeline.","storageIdentifier":"file://19ef0982a26-b8a2aa65b84e","rootDataFileId":-1,"md5":"050192c781468337fb7718810e8316b7","checksum":{"type":"MD5","value":"050192c781468337fb7718810e8316b7"},"creationDate":"2026-06-22"}}],"citation":"Zuzana Černáková; Fien Messens; Tess Dejaeghere; Julie M. Birkholz, 2026, \"Replication Data for: From nineteenth-century letters to entities: \"a NER pipeline for French correspondence and its methodological lessons\" - Article for Digital Humanities Benelux Journal\", https://doi.org/10.34934/DVN/HNA7QO, Social Sciences and Digital Humanities Archive – SODHA, V1, UNF:6:MemhfA5Fl4s1tvLWf4VR2A== [fileUNF]"}}