Rhyme, Syllabic, and Repetition Dataset of English Nursery Rhymes (RSR)

Description

The Rhyme, Syllabic, and Repetition Dataset of English Nursery Rhymes (RSR) provides document-level structural measurements for 308 English nursery rhymes. Each nursery rhyme is represented by twelve structural features covering three linguistic dimensions: rhyme, syllabic structure, and repetition. Rhyme is represented by Rhyme Participation Rate, Rhyme Density, Number of Rhyme Classes, Perfect Rhyme Lines, Perfect Rhyme Pairs, and Rhyme Scheme. Syllabic structure is represented by Mean Syllable Count, SD Syllable Count, and Syllabic CV. Repetition is represented by Word Repetition Rate, Phrase Repetition Rate, and Line Repetition Rate. The repository also includes line-level processing records, pronunciation and out-of-vocabulary audits, syllable-estimation records, repetition audits, consistency checks, and supporting files for dataset validation and computational reproducibility. The dataset is intended to support structural analysis of English nursery rhymes and may be used for comparative analysis, clustering, classification, and other computational linguistics or natural language processing applications.

Steps to reproduce

1. Obtain the 308 English nursery rhyme source text files from the original Kaggle distribution: English Nursery Rhymes by Terence Broad. 2. Run Final Dataset/nursery_rhyme_structural_dataset.py on the defined 308-file source corpus. The script reconstructs the document-level structural dataset and generates the XLSX, CSV, and JSON outputs. It also creates a local supporting workbook named validation_02_supporting_processing.xlsx for computational consistency validation. 3. Run Validation/01 Dataset Integrity and Completeness/validation_01_integrity_completeness.py using the reconstructed public Main Dataset. This validation checks record completeness, identifiers, missing values, duplicates, and availability of the twelve structural features. 4. Run Validation/02 Computational Consistency/validation_02_computational_consistency.py using the validation_02_supporting_processing.xlsx file produced during reconstruction. This validation recalculates the twelve structural features from the supporting processing records and compares them with the published feature table. 5. Run the dataset construction pipeline independently a second time using the same defined 308-file source corpus. 6. Run Validation/03 Computational Reproducibility/validation_03_computational_reproducibility.py to compare the deposited reference workbook with the workbook generated by the independent rerun. Reproducibility is evaluated through record-level and feature-level comparisons.

Institutions

• State University of Malang

  East Java, Malang

Categories

Licence