Codebase and issue tracking


childes-db is all open source and is under active development. If you run into an issue, please let us know by github's issue tracking at the appropriate repository:

Website frontend: http://github.com/langcog/childes-db-website
Interactive applications: https://github.com/langcog/childes-db-shiny
R package: https://github.com/langcog/childesr

If you have any other questions about childes-db or want to stay updated on changes to the database, visualizations, or API, join our mailing list. If you would like to contact us directly, email us at childes-db-contact@stanford.edu.


Backend and CHAT format treatment


The childes-db project stores CHILDES data in a MySQL relational database. The original CHAT files were converted to XML via chatter, a cross-platform application provided by TalkBank. The XML files were then parsed using the Python xml library and subsequently imported into a MySQL database through Django. The annotations and markers from the original CHAT files that were preserved in childes-db can be found in the table below:




Database versioning


One design goal for childes-db is to ensure reproducibility in analyses. The database is versioned and both the API and visualizations provide the database version for each analysis - we recommend that this detail be reported with any analyses. To reproduce an earlier version of the database, we currently provide direct downloads below, and we hope in the future to make these previous versions more directly accessible through the API. For additional information regarding the database schema, sample queries and instructions regarding local installation of the database, refer to the documentation for childes-db. Note that API access should handle most use cases for most users.