french_first_names_insee_2024

# French First Names from Death Records (1970-2024) This dataset contains French first names extracted from death records provided by INSEE (French National Institute of Statistics and Economic Studies) covering the period from 1970 to September 2024. ## Dataset Description ### Data Source The data is sourced from INSEE's death records database. It includes first names of deceased individuals in France, providing valuable insights into naming patterns across different generations. The extractor source code can be found on Github [sctg-development/french-names-extractor](https://github.com/sctg-development/french-names-extractor) ### Time Period - Start: 1970 - End: September 2024 ### Features The dataset contains three columns: - `firstname`: The first name of the deceased person - `sexe`: Gender indicator (1 = male, 2 = female) - `occurrences`: Number of occurrences of the name in the records ### Data Processing The following filtering criteria were applied to clean the data: - Names with length > 1 character - Names not composed of repeated single characters - Only names with at least 2 occurrences are included ### Data Format - File format: CSV - Encoding: UTF-8 - Separator: comma (,) ## Usage This dataset can be used for various purposes, including: - Historical analysis of French naming patterns - Demographic studies - Gender classification of French names - Cultural evolution studies - Genealogical research ### Loading the Dataset ```python from datasets import load_dataset dataset = load_dataset("eltorio/french_first_names_insee_2024") ``` ## License MIT License ## Citation ```bibtex @dataset{french_first_names_insee_2024, author = {Ronan Le Meillat}, title = {French First Names from Death Records (1970-2024)}, year = {2024}, publisher = {Hugging Face}, source = {INSEE}, } ``` ## Dataset Creation ### Data Collection The data was extracted from INSEE's death records database, which maintains official records of all deaths in France. ### Preprocessing 1. Extraction of first names from death records 2. Filtering out names with only 1 character 3. Filtering out names composed of repeated single characters 4. Removal of names with single occurrence 5. Gender coding (1 for male, 2 for female) 6. Counting occurrences of each name ## Dataset Structure ```python DatasetDict({ 'train': Dataset({ features: ['firstname', 'sexe', 'occurrences'], num_rows: [NUMBER_OF_ROWS] }) }) ``` ## Additional Information ### Source Data The original data was obtained from INSEE (Institut National de la Statistique et des Études Économiques), the French National Institute of Statistics and Economic Studies. ### Maintenance This dataset is a static snapshot and reflects death records up to September 2024. ### Known Limitations - The dataset only includes names from death records, which may not fully represent contemporary naming patterns - Historical records might have some inconsistencies in spelling or recording - The dataset only includes names with at least 2 occurrences, which might exclude some rare or unique names ## Acknowledgments Thanks to INSEE for making this data publicly available.