Carepart/repairers-france
收藏Hugging Face2026-04-12 更新2026-04-26 收录
下载链接:
https://hf-mirror.com/datasets/Carepart/repairers-france
下载链接
链接失效反馈官方服务:
资源简介:
---
license: cc-by-sa-4.0
language:
- fr
tags:
- france
- repair
- circular-economy
- sustainability
- directory
- business
- local-business
- eco-design
size_categories:
- 10K<n<100K
task_categories:
- text-retrieval
- question-answering
pretty_name: "Carepart.fr — French Repairers Directory"
configs:
- config_name: default
data_files:
- split: repairers
path: "data/repairers.parquet"
- split: categories
path: "data/categories.parquet"
- split: city_category_stats
path: "data/category_city_stats.parquet"
- split: indice_reparabilite_ademe
path: "data/indice_reparabilite_ademe.parquet"
---
# Carepart.fr — French Repairers Directory
> The most comprehensive open dataset of repair professionals in France: 38 000+ active repairers across 95 categories of objects, with QualiRépar / Refashion certification status, Google reviews, geolocation, and AI-generated contextual descriptions for each business.
[](https://carepart.fr)
[](https://creativecommons.org/licenses/by-sa/4.0/)
## Dataset Summary
This dataset is a snapshot of the **Carepart.fr** repair professionals directory — a French-language platform that helps consumers find local repairers and benefit from the **Bonus Réparation** (a state-funded incentive of up to €60 to encourage repair over replacement).
It is intended for:
- **Research** on the circular economy, repair access, and sustainability in France
- **NLP / RAG / LLM training** : grounded answers to questions about repair services, local businesses, eco-friendly consumption
- **Geographic analysis** : repair coverage by department, urban/rural divide, label distribution
- **Civic tech** : building tools that help consumers reduce waste
## Data Splits
| Split | Rows | Description |
|---|---|---|
| `repairers` | ~38 000 | Active repair businesses with categories, location, ratings, AI descriptions |
| `categories` | 95 | Repair categories (smartphone, washing machine, bike, furniture, etc.) with bonus amounts and lifecycle data |
| `city_category_stats` | ~190 000 | Counts of repairers per (category, city) combo, label coverage, average rating |
| `indice_reparabilite_ademe` | ~3 100 | Repairability and durability index of consumer products from the French ADEME |
## Schema
### `repairers`
| Column | Type | Description |
|---|---|---|
| `id` | int | Unique internal ID |
| `slug` | string | URL-friendly identifier |
| `siret` | string | French business identifier (SIRENE database) |
| `siren` | string | Parent legal entity ID |
| `name` | string | Trade name or company name |
| `legal_form` | string | Legal status (SARL, micro-entreprise, etc.) |
| `naf_code` | string | French business activity code |
| `naf_label` | string | Activity description |
| `address` | string | Street address |
| `postal_code` | string | French postal code |
| `city` | string | City name |
| `city_code_insee` | string | Official INSEE city code |
| `city_slug` | string | URL-friendly city identifier (e.g. `paris-75`) |
| `department_code` | string | French department code (e.g. `75`, `2A`) |
| `latitude` | float | WGS84 latitude (6 decimals) |
| `longitude` | float | WGS84 longitude |
| `phone` | string | Phone number (only if business landline / professional) |
| `email` | string | Email (only if business domain — personal emails removed for privacy) |
| `website` | string | Business website |
| `is_qualirepar` | bool | Certified for the QualiRépar Bonus Réparation |
| `is_refashion` | bool | Certified for the Refashion (textile) Bonus Réparation |
| `qualirepar_id` | string | Official QualiRépar identifier |
| `google_rating` | float | Average Google review rating (1.0–5.0) |
| `google_review_count` | int | Number of Google reviews |
| `repair_modes` | list[string] | Service modes: `boutique`, `domicile`, `postal`, etc. |
| `category_slugs` | list[string] | Repair categories handled |
| `category_names` | list[string] | Human-readable category names |
| `ai_description` | string | AI-generated narrative describing the repairer in its local context (~200 words) |
| `page_url` | string | Canonical URL on carepart.fr |
### `categories`
| Column | Type | Description |
|---|---|---|
| `id` | int | |
| `slug` | string | URL-friendly slug (e.g. `reparation-smartphone`) |
| `name` | string | Display name |
| `name_plural` | string | Plural form |
| `parent_group` | string | Parent category group |
| `parent_group_label` | string | Display label of the parent group |
| `bonus_amount` | float | Maximum Bonus Réparation amount in euros |
| `bonus_source` | string | Source of the bonus (`qualirepar`, `refashion`) |
| `icon_emoji` | string | Display emoji |
| `description` | string | Category description |
| `avg_lifespan` | string | Average lifetime of the product type |
| `avg_price_min/max` | float | Average repair price range |
| `new_price_min/max` | float | Average new product price range |
| `co2_saved_kg` | float | Estimated CO₂ saved per repair |
| `waste_saved_kg` | float | Estimated waste saved per repair |
### `city_category_stats`
| Column | Type | Description |
|---|---|---|
| `category_id`, `category_slug` | int / string | |
| `city_code_insee`, `city_slug`, `city_name` | string | |
| `dept_code` | string | French department code |
| `repairer_count` | int | Total active repairers in this combo |
| `qualirepar_count` | int | Of which QualiRépar-certified |
| `refashion_count` | int | Of which Refashion-certified |
| `avg_google_rating` | float | Average Google rating across listed repairers |
### `indice_reparabilite_ademe`
Repairability and durability scores from the French [ADEME](https://www.ademe.fr/) (data.gouv.fr).
| Column | Type | Description |
|---|---|---|
| `categorie_produit` | string | Product category |
| `marque`, `nom_modele` | string | Brand and model |
| `note_ir` | float | Repairability score (0–10) |
| `note_durabilite` | float | Durability score |
| `note_fiabilite` | float | Reliability score |
| `note_c1`–`note_c5` | float | Sub-criteria (documentation, dismantling, parts availability, parts pricing, specific to category) |
| `type_indice` | string | `reparabilite` or `durabilite` |
| `date_calcul` | date | Calculation date |
| `url_detail` | string | Manufacturer documentation URL |
| `data_source` | string | Origin (data.gouv.fr) |
## Sources
This dataset aggregates several public sources :
- **[SIRENE](https://www.data.gouv.fr/fr/datasets/base-sirene-des-entreprises-et-de-leurs-etablissements-siren-siret/)** (INSEE) — French business registry, for legal information (SIRET, address, NAF code).
- **[Annuaire QualiRépar](https://www.ecosystem.eco/reparer/)** (ecosystem.eco) — list of certified repairers eligible for the electronic/appliance Bonus Réparation.
- **[Refashion](https://refashion.fr/citoyen/fr/bonus-reparation)** — list of certified textile and shoe repairers.
- **[Google Business Profile](https://www.google.com/business/)** — public reviews, ratings, opening hours.
- **[Indice de réparabilité (ADEME)](https://www.data.gouv.fr/fr/datasets/indice-de-reparabilite/)** — official French repairability scores.
- **AI-generated descriptions** — original narrative content produced by the Carepart.fr team using Claude (Anthropic), released under the same CC-BY-SA-4.0 license as the rest of the dataset.
## Privacy & Anonymization
- Personal email addresses (gmail, hotmail, free.fr, orange.fr, etc.) have been **stripped**. Only emails on a business domain are kept.
- Phone numbers are kept (already public via SIRENE).
- No personal customer data, no transactional data, no PII beyond what is publicly available.
## Example Usage
### Load with `datasets`
```python
from datasets import load_dataset
ds = load_dataset("carepart/repairers-france")
# Top 10 QualiRépar bike repairers in Paris
import pandas as pd
df = pd.DataFrame(ds["repairers"])
paris_velo = df[
(df["category_slugs"].apply(lambda c: "reparation-velo" in c)) &
(df["department_code"] == "75") &
(df["is_qualirepar"] == True)
].sort_values("google_rating", ascending=False).head(10)
print(paris_velo[["name", "address", "google_rating"]])
```
### Load with `pandas`
```python
import pandas as pd
repairers = pd.read_parquet("hf://datasets/carepart/repairers-france/data/repairers.parquet")
stats = pd.read_parquet("hf://datasets/carepart/repairers-france/data/category_city_stats.parquet")
# Departments with the most QualiRépar coverage
top_dept = (
repairers.groupby("department_code")["is_qualirepar"]
.agg(["sum", "count"])
.assign(rate=lambda d: (d["sum"] / d["count"] * 100).round(1))
.sort_values("rate", ascending=False)
.head(15)
)
```
### RAG / LLM grounding
Each repairer entry includes an `ai_description` field — a ~200-word narrative describing the repair professional in its local context (city, specialties, sustainability angle). This is a perfect substrate for retrieval-augmented generation systems answering questions about local repair services in France.
## License
- **Dataset & AI descriptions**: [Creative Commons Attribution-ShareAlike 4.0 International (CC-BY-SA 4.0)](https://creativecommons.org/licenses/by-sa/4.0/)
- **Original sources** retain their respective licenses (SIRENE Open License, ADEME Open License, Google ToS for derived metrics).
## Citation
If you use this dataset in research or commercial work, please cite :
```bibtex
@dataset{carepart_repairers_2026,
title = {Carepart.fr — French Repairers Directory},
author = {Charpentier, Eric and the Carepart.fr team},
year = {2026},
url = {https://huggingface.co/datasets/carepart/repairers-france},
publisher = {Carepart.fr},
license = {CC-BY-SA-4.0},
note = {Aggregated from SIRENE, ecosystem.eco, Refashion, ADEME, and Google Business Profile.},
}
```
## Contact & Updates
- **Live website**: <https://carepart.fr>
- **Email**: bonjour@carepart.fr
- **Updates**: this dataset is refreshed periodically. Check the export date in `data/_summary.json`.
## Acknowledgements
Thanks to the French circular-economy ecosystem — **ADEME**, **ecosystem.eco**, **Refashion**, and the thousands of independent repair artisans across France whose work makes this directory meaningful.
提供机构:
Carepart



