This project represents a real world project built to map attribute data between two existing product systems. The goal was to migrate data from two older existing systems into a new taxonomy structure with the attribute mapping serving as a guide for how product data will be represented in the new system.
The project used NLP to suggest potential attribute mappings based on the similarities of top values in product attribute values within similar taxonomy nodes. These mappings were manually reviewed by the taxonomy team and we are current in the process of making the required moves.
Because this project is public, that data used here is an artifically constrained sample and not representative of actual product data.