# Processed Provo Corpus — Licensing, Attribution, and Citation **Repository:** AffLex-Provo **Research Program:** AffLex-Provo **Computational Processing and Development:** Mohit Gupta **Institution:** University of Illinois Urbana-Champaign **Year:** 2026 **Repository URL:** https://huggingface.co/datasets/mohitg2/AffLex-Provo --- ## 1. Overview This directory contains intermediate research datasets generated through computational preprocessing of the Provo Corpus as part of the AffLex-Provo research program. The datasets were processed and prepared by Mohit Gupta using computational procedures developed for the project. The processing workflow includes data cleaning, standardization, restructuring, validation, preparation of predictability information, vocabulary matching, and lemmatization-related analyses. These intermediate datasets support subsequent integration of lexical, affective, and eye-movement information for psycholinguistic research. The datasets contain information derived from previously published research resources. Accordingly, original computational contributions and underlying third-party data are distinguished throughout this document. --- ## 2. Dataset Inventory This directory contains four processed or derived datasets. | File | Description | Contribution | |---|---|---| | `1_Provo_Norming_Item_Level_Cloze.csv` | Processed item-level cloze predictability data derived from the Provo Corpus. | Data preparation, restructuring, and validation. | | `2_processed_Provo_Corpus.csv` | Processed token-level Provo Corpus prepared for statistical analysis and subsequent data integration. | Computational preprocessing, standardization, and validation. | | `3_provo_missing_vocabulary_pre_lemmatization.csv` | Vocabulary coverage report identifying unmatched words before lemmatization. | Computational vocabulary matching and identification of coverage gaps. | | `4_provo_missing_vocabulary_post_lemmatization.csv` | Vocabulary coverage report identifying unmatched words after lemmatization. | Lemmatization-related processing and vocabulary coverage analysis. | All four files were processed or generated through the AffLex-Provo computational research workflow. The original Provo Corpus remains attributable to its original creators. --- ## 3. Original Computational Contributions The AffLex-Provo research workflow includes computational contributions developed by Mohit Gupta. These contributions include, where applicable: 1. Cleaning and standardization of source data. 2. Restructuring of item-level predictability information. 3. Preparation of token-level eye-tracking observations. 4. Validation of processed research datasets. 5. Vocabulary extraction and matching. 6. Identification of missing lexical entries. 7. Lemmatization-related vocabulary processing. 8. Generation of intermediate research datasets. 9. Preparation of data for subsequent lexical and affective integration. These contributions are documented through the corresponding research code and preprocessing notebooks. The processing workflow does not imply ownership of the original Provo Corpus or its underlying observations. --- ## 4. Licensing of Original Contributions ### 4.1 Creative Commons Attribution 4.0 International To the extent that the AffLex-Provo contributors possess the necessary intellectual property rights, eligible original contributions to the intermediate datasets are made available under: **Creative Commons Attribution 4.0 International (CC BY 4.0)** Official license: https://creativecommons.org/licenses/by/4.0/ Under this license, users may: - Copy and redistribute licensed materials. - Share licensed materials with other researchers. - Adapt, transform, and build upon licensed materials. - Incorporate licensed materials into new research. - Use licensed materials for commercial or noncommercial purposes. Users must comply with the applicable conditions of CC BY 4.0, including appropriate attribution, indicating modifications, and providing a link to the license. The license applies only to contributions for which the AffLex-Provo contributors hold sufficient rights. It does not override third-party intellectual property rights or redistribution restrictions. ### 4.2 Scope of the License The CC BY 4.0 license applies to eligible original contributions developed through the AffLex-Provo research workflow. These may include original documentation, annotations, and other protectable contributions incorporated into the intermediate research materials. The license does not automatically apply to: - Original Provo Corpus observations. - Original eye-tracking measurements. - Original cloze predictability norms. - Third-party lexical frequency data. - Third-party affective norming data. - Materials governed by incompatible source licenses. - Contributions for which the project contributors do not possess the necessary licensing rights. Mechanical processing, filtering, or restructuring does not automatically create new copyright protection in underlying factual data. --- ## 5. Original Provo Corpus Attribution The intermediate datasets are derived primarily from the Provo Corpus. ### Original Publication Luke, S. G., & Christianson, K. (2018). The Provo Corpus: A large eye-tracking corpus with predictability norms. *Behavior Research Methods, 50*, 826–833. https://doi.org/10.3758/s13428-017-0908-4 ### Original Data Rights The Provo Corpus was developed by its original research team. Its underlying data remain subject to the applicable rights, licenses, permissions, and distribution conditions established by the relevant rights holders. The AffLex-Provo project does not claim exclusive ownership of the original Provo Corpus. Processing, cleaning, restructuring, or transforming the original data does not automatically transfer ownership of the underlying source material. Public redistribution of source-derived datasets must comply with the applicable original permissions. Researchers using Provo-derived information should cite Luke and Christianson (2018). --- ## 6. Scholarly Citation of AffLex-Provo Contributions ### 6.1 Requested Citation **If you use the original AffLex-Provo preprocessing contributions, intermediate datasets, or associated computational research resources in academic work, please cite the corresponding AffLex-Provo publication.** ### Associated Research Publication Gupta, M., & Christianson, K. (in press). Beyond word length, frequency, and predictability: Affective properties predict eye movement in reading. *Reading Research Quarterly.* **Publication status:** Accepted October 20, 2026. **DOI:** Pending assignment. The final citation and DOI will be updated when the article becomes available online. ### 6.2 Citation Expectations Researchers are encouraged to cite the associated AffLex-Provo publication when using original contributions from this repository in: - Peer-reviewed research articles. - Conference papers and presentations. - Dissertations and theses. - Research reports. - Computational research projects. - Dataset integration studies. - Reproducibility and replication studies. - Methodological research. Researchers should also cite the original publications associated with any third-party datasets they use. ### 6.3 Attribution Versus Academic Citation CC BY 4.0 requires appropriate attribution of licensed material. However, its attribution requirements do not necessarily require a formal bibliographic citation of the associated research article. The request to cite the AffLex-Provo publication is a scholarly citation expectation and is separate from the legally enforceable conditions of CC BY 4.0. This document does not impose an additional publication-citation condition beyond the standard CC BY 4.0 license. --- ## 7. Recommended Attribution Statement When using original AffLex-Provo contributions, researchers are encouraged to acknowledge the processing work using a statement such as: > The intermediate research data were prepared using > computational preprocessing resources developed > by Mohit Gupta as part of the AffLex-Provo research > program at the University of Illinois > Urbana-Champaign. When the underlying Provo Corpus is used, researchers should additionally acknowledge its original creators. Suggested acknowledgment: > The research data were derived from the Provo Corpus > (Luke & Christianson, 2018) and processed using > computational resources developed for the > AffLex-Provo research program > (Gupta & Christianson, in press). These statements are suggested acknowledgments, not mandatory wording. --- ## 8. Redistribution and Derivative Works Users may redistribute or adapt eligible original AffLex-Provo contributions in accordance with their applicable licenses. When redistributing or adapting licensed materials, users must comply with the relevant license conditions. For CC BY 4.0 materials, these include: 1. Providing appropriate attribution. 2. Providing a link to the CC BY 4.0 license. 3. Indicating whether modifications were made. 4. Not implying endorsement by the original creators. 5. Not imposing additional restrictions that conflict with the license. Third-party materials remain subject to their respective original terms. Users must independently ensure that any redistribution of source-derived data is authorized. The CC BY 4.0 license in this document does not constitute permission to redistribute third-party materials where such permission has not been granted. --- ## 9. Research Reproducibility The intermediate datasets were generated through the AffLex-Provo computational preprocessing pipeline. The corresponding preprocessing scripts and notebooks document the procedures used to prepare the datasets. Researchers are encouraged to consult these materials when reproducing or extending the processing workflow. The availability of processed datasets does not eliminate the need to acknowledge the original research resources from which they were derived. Where source-data redistribution is restricted, researchers should obtain the original materials through their authorized distribution channels. --- ## 10. Additional Source Citations The broader AffLex-Provo research program also uses lexical frequency and affective norming resources. These include: - SUBTLEX-US. - Warriner et al. (2013) affective norms. The corresponding original publications and attribution information are documented in: [Mandatory Citations](../01-Raw/MandatoryCitations.md) Researchers should cite the source publications relevant to the specific data used in their work. --- ## 11. Citation and Licensing Updates This document may be updated to reflect: - Assignment of the AffLex-Provo publication DOI. - Publication of additional associated research papers. - Verification of third-party licensing conditions. - Changes to documentation. - Release of additional processed datasets. - Identification of specific dataset versions. Any updates to licensing statements will respect previously granted permissions and applicable third-party rights. --- ## 12. Repository Information **Repository:** AffLex-Provo **Research Program:** AffLex-Provo **Computational Processing:** Mohit Gupta **Institution:** University of Illinois Urbana-Champaign **Repository URL:** https://huggingface.co/datasets/mohitg2/AffLex-Provo **Associated Journal:** Reading Research Quarterly **Last Updated:** October 2026 --- ## 13. Important Legal Notice This document identifies the intended licensing and attribution arrangements for original AffLex-Provo contributions. It does not establish ownership of third-party datasets or grant permissions beyond those held by the applicable rights holders. The presence of a dataset in this repository does not necessarily indicate that all its contents are independently licensed under CC BY 4.0. Users must comply with all applicable third-party licenses and permissions.