charity-classification/ukcat

Implements dual classification approaches: rule-based regex matching for UK-CAT categories and logistic regression ML models for ICNP/TSO taxonomy, both trained on manually-classified charity samples. Aggregates charity data from three UK regulatory bodies (England & Wales, Scotland, Northern Ireland) and applies classifications via a Python CLI module (`ukcat`). Documentation and Jupyter notebooks in the repo support reproducible analysis, with schema and results published under CC-BY 4.0 licensing.