| { | |
| "normalization": "ncm_classifier.text.normalize.normalize_text", | |
| "applied_by": "both TF-IDF vectorizers of the pipeline (preprocessor)", | |
| "steps": [ | |
| "Unicode NFKC, lowercase, accents removed", | |
| "fertilizer formulas standardized (NPK 20 05 20 -> npk 20-05-20)", | |
| "decimal commas, units and abbreviations standardized", | |
| "punctuation and extra spaces removed" | |
| ], | |
| "rule_31_05": { | |
| "module": "ncm_classifier.rules.small_package", | |
| "applied_after": "calibration", | |
| "limit_kg": 10.0, | |
| "target": "3105" | |
| }, | |
| "package": "ncm-classifier @ git+https://github.com/amanda-gondra/ncm-classifier@v0.1.0" | |
| } | |