← back to Handbag Authentication

INTERNATIONAL_HANDBAG_DATA_SOURCES.md

817 lines

# International Handbag Data Sources
## Focus: French & European Luxury Brands

**Generated:** 2025-11-11
**Geographic Coverage:** France, Italy, Netherlands, Japan, Pan-European

---

## 🇫🇷 French Sources - Luxury Brand Focus

### 1. FANCY Dataset (Fashion Show Archive) ⭐⭐⭐⭐⭐

**Repository:** https://github.com/youngseungjeon/FANCY

**Coverage:**
- **Total Images:** 302,772 runway images
- **Fashion Shows:** 8,121 shows from Vogue US (2010-2019)
- **Brands:** 987 brands including French luxury houses
- **Attributes:** 146 fashion attributes
- **Style Groups:** 25 categories

**French Brands Included:**
- Hermès ✅
- Chanel ✅
- Louis Vuitton ✅
- Dior ✅
- Celine ✅
- Balenciaga ✅
- Yves Saint Laurent ✅
- Givenchy ✅
- Longchamp ✅

**Data Access:**
- ADM Dataset: https://drive.google.com/drive/folders/1rvd5D0vv2MrMEaZZpmeJ2ny6Tn44DyA-
- FANCY Dataset: https://drive.google.com/drive/folders/1abIiasmgCSdvNpEDJM9-2iaURlTQh8PY

**License:** Non-commercial research only

**Repository Status:** ✅ Cloned to `handbag_data/github_datasets/FANCY/`

**Data Quality:** Excellent - Professional runway photography with multi-label annotations

---

### 2. Europeana Fashion Collections 🇪🇺

**API:** https://api.europeana.eu/
**Website:** https://www.europeana.eu/

**Coverage:**
- **Total Items:** 50 million+ cultural artifacts
- **Fashion Specific:** Extensive European fashion archives
- **Key Partners:**
  - Palais Galliera (Paris Fashion Museum)
  - Musée des Arts Décoratifs (Paris)
  - French regional museums

**French Content:**
- Palais Galliera collection (200,000+ fashion items)
- French National Library fashion documents
- Regional French museum collections
- Historical fashion illustrations

**Search Terms (French):**
- `sac à main` - handbag
- `maroquinerie` - leather goods
- `Hermès`, `Chanel`, `Louis Vuitton`
- `haute couture`
- `mode française` - French fashion

**API Access:**
- Registration: https://pro.europeana.eu/page/get-api
- Free API key required
- Rate limit: Reasonable for research

**Query Example:**
```
https://api.europeana.eu/record/v2/search.json?
wskey=YOUR_KEY
&query=sac à main
&qf=PROVIDER:"Europeana Fashion"
&rows=100
```

**Data Format:** JSON with multilingual metadata

**Status:** ⏳ Requires API key registration

---

### 3. Gallica - Bibliothèque nationale de France (BNF)

**Website:** https://gallica.bnf.fr/
**API:** https://api.bnf.fr/

**Coverage:**
- **Total Items:** Millions of digitized documents
- **Fashion Content:** Historical fashion plates, magazines, catalogs
- **Time Period:** 18th century to present
- **Language:** Primarily French

**Fashion Resources:**
- Historical fashion magazines
- Designer sketches and catalogs
- Advertising materials
- Fashion photography archives
- Hermès, Chanel historical documents

**Search Terms:**
- `sac Hermès`
- `maroquinerie de luxe`
- `mode parisienne`
- `accessoires de mode`
- `catalogue Louis Vuitton`

**Access Methods:**
1. **Web Search:** https://gallica.bnf.fr/accueil/en/
2. **API:** IIIF Image API + Search API
3. **Bulk Download:** Available for large collections

**Data Types:**
- Images (IIIF format)
- Text (OCR available)
- Metadata (MARC, Dublin Core)

**Status:** ✅ Publicly accessible, no API key for basic access

**Note:** Primarily historical/archival rather than contemporary product data

---

### 4. Palais Galliera (Paris Fashion Museum)

**Website:** https://www.palaisgalliera.paris.fr/en
**Via Europeana:** Available through Europeana API

**Collection Size:** 200,000+ works

**Accessories Collection:**
- Handbags and purses ✅
- Jewelry
- Gloves
- Scarves
- Hats
- Fans

**Historical Coverage:** 18th century to present

**Key Holdings:**
- French designer pieces
- Hermès archival pieces
- Chanel accessories
- Historical Parisian fashion

**Access:**
1. **Europeana API** (primary method)
2. **Paris Musées API:** https://apicollections.parismusees.paris.fr/
3. **Direct website:** Digital resources and virtual tours

**Data Format:**
- Via Europeana: JSON
- Paris Musées: GraphQL API

**Status:** ⏳ Accessible via Europeana API

---

### 5. Louis Vuitton Web Scraped Data

**Repository:** https://github.com/databoutique/Louis-Vuitton-web-scraped-data

**Content:** Product catalog snapshot

**Data Type:** Web scraping dataset (commercial)

**Coverage:**
- Current product line
- Pricing information
- Product images
- Product specifications

**Status:** ✅ Cloned to `handbag_data/github_datasets/Louis-Vuitton-web-scraped-data/`

**Note:** README only, actual data may require purchase from Databoutique.com

**Competitors Mentioned:** Prada, Chanel, Gucci

---

## 🇮🇹 Italian Sources

### 6. Gucci Web Scraped Data

**Repository:** https://github.com/databoutique/Gucci-web-scraped-data

**Content:** Product catalog snapshot

**Coverage:**
- Current Gucci product line
- Italian fashion leader
- Complete website snapshot

**Status:** ✅ Cloned to `handbag_data/github_datasets/Gucci-web-scraped-data/`

**Note:** README only, actual data commercial

---

### 7. Italian Fashion in FANCY Dataset

**Brands Covered:**
- Gucci ✅
- Prada ✅
- Fendi ✅
- Versace ✅
- Bottega Veneta ✅
- Salvatore Ferragamo ✅
- Valentino ✅
- Dolce & Gabbana ✅

**Access:** Via FANCY dataset (see above)

**Coverage:** 302,772 fashion show images (2010-2019)

---

## 🇳🇱 Netherlands Sources

### 8. Rijksmuseum Amsterdam

**Website:** https://www.rijksmuseum.nl/
**API:** https://data.rijksmuseum.nl/

**Coverage:**
- **Total Objects:** 500,000+
- **With Images:** 700,000+ images
- **Fashion & Textiles:** Extensive collection

**Fashion Collection:**
- Historical bags and purses
- Dutch fashion history
- International designer pieces
- Textile arts

**Search Terms (Dutch):**
- `handtas` - handbag
- `tas` - bag
- `beurs` - purse
- `mode` - fashion

**API Access:**
- Free registration required
- Documentation: https://data.rijksmuseum.nl/object-metadata/api/
- Format: JSON
- License: CC0 for metadata

**Query Example:**
```
https://www.rijksmuseum.nl/api/nl/collection?
key=YOUR_KEY
&q=handtas
&imgonly=true
&ps=100
```

**Status:** ⏳ Requires free API key

---

### 9. Museum of Bags and Purses (Tassenmuseum)

**Status:** ⚠️ CLOSED (April 2020 due to COVID-19)

**Historical Note:**
- Was first museum dedicated to bags/purses
- Collection fate: Unknown/redistributed
- Historical data may be in Rijksmuseum or other Dutch institutions

---

## 🇯🇵 Japanese Sources

### 10. Kyoto Costume Institute (KCI)

**Website:** https://www.kci.or.jp/en/
**Digital Archives:** https://www.kci.or.jp/en/archives/digital_archives/

**Collection Size:**
- **Total:** 13,000 clothing items + 20,000 documents
- **Online Public Access:** 300 selected items
- **Full Access:** By reservation (Study Room)

**Time Period:** 17th century to present

**Coverage:**
- Western haute couture
- Japanese fashion
- Designer pieces from major houses

**French Brands in Collection:**
- Chanel ✅
- Hermès ✅
- Balenciaga ✅
- Other French couturiers

**Access Methods:**
1. **Online:** 300 items publicly viewable
2. **On-site Study Room:** Full 13,000+ item access (by appointment)
3. **Publications:** Fashion history books

**Data Format:**
- Images (high quality)
- Bilingual (Japanese/English)
- Detailed provenance

**Search Terms (Japanese):**
- `ハンドバッグ` (handobaggu) - handbag
- `バッグ` (baggu) - bag
- `財布` (saifu) - purse/wallet

**Status:** ✅ 300 items publicly accessible online

**URL:** https://www.kci.or.jp/en/archives/digital_archives/

---

## 🌍 Pan-European Sources

### 11. Paris Musées Collections API

**Website:** https://www.parismusees.paris.fr/
**API:** https://apicollections.parismusees.paris.fr/

**Coverage:**
- **Total Works:** 280,000+
- **Museums:** 14 Paris municipal museums
- **Includes:** Palais Galliera fashion collections

**API Type:** GraphQL

**Access:** Open API, no key required

**Data License:** Open content (150,000+ works free to use)

**Fashion Content:**
- Palais Galliera items
- Decorative arts
- Historical textiles

**Status:** ✅ Public API available

**Documentation:** https://apicollections.parismusees.paris.fr/en

---

### 12. Victoria & Albert Museum (Already Covered)

**Location:** London, UK

**French/European Content:**
- French designer pieces
- European fashion history
- International luxury brands

**Status:** ✅ 300 items fetched (see main report)

---

## 📊 Data Summary by Country

### France 🇫🇷
| Source | Items | Type | Status |
|--------|-------|------|--------|
| FANCY Dataset | 302,772 | Images (shows) | ✅ Ready to download |
| Europeana Fashion | 1M+ | Museum items | ⏳ API key needed |
| Gallica BNF | Millions | Archives | ✅ Accessible |
| Palais Galliera | 200K+ | Museum | ⏳ Via Europeana |
| Louis Vuitton Data | Unknown | Products | ⏳ Commercial |

### Italy 🇮🇹
| Source | Items | Type | Status |
|--------|-------|------|--------|
| FANCY Dataset | Included | Images | ✅ Ready to download |
| Gucci Data | Unknown | Products | ⏳ Commercial |
| Europeana | Varies | Museums | ⏳ API key needed |

### Netherlands 🇳🇱
| Source | Items | Type | Status |
|--------|-------|------|--------|
| Rijksmuseum | 500K+ | Museum | ⏳ API key needed |

### Japan 🇯🇵
| Source | Items | Type | Status |
|--------|-------|------|--------|
| KCI Kyoto | 300/13K | Museum | ✅ 300 online |

### EU-Wide 🇪🇺
| Source | Items | Type | Status |
|--------|-------|------|--------|
| Europeana | 50M+ | Pan-EU | ⏳ API key needed |

---

## 🔤 Translation Guide

### French Search Terms
```
English → French
-------------------
handbag → sac à main
purse → bourse / porte-monnaie
bag → sac
clutch → pochette
tote bag → cabas
shoulder bag → sac porté épaule
crossbody bag → sac bandoulière
leather goods → maroquinerie
luxury → luxe
designer → créateur / styliste
fashion → mode
vintage → vintage
collection → collection
```

### Italian Search Terms
```
English → Italian
-------------------
handbag → borsetta
purse → borsellino
bag → borsa
leather goods → pelletteria
luxury → lusso
fashion → moda
designer → stilista
```

### Dutch Search Terms
```
English → Dutch
-------------------
handbag → handtas
purse → beurs
bag → tas
fashion → mode
leather → leer
```

### Japanese Search Terms
```
English → Japanese (Romaji)
-------------------
handbag → ハンドバッグ (handobaggu)
bag → バッグ (baggu)
purse → 財布 (saifu)
fashion → ファッション (fasshon)
luxury → 高級 (kōkyū)
brand → ブランド (burando)
```

---

## 🔧 Technical Integration

### Translation Script Setup

```javascript
const translations = {
  handbag: {
    fr: 'sac à main',
    it: 'borsetta',
    nl: 'handtas',
    ja: 'ハンドバッグ',
    de: 'Handtasche',
    es: 'bolso'
  },
  // ... more terms
};

function searchMultilingual(term, apis) {
  const queries = Object.keys(translations[term]).map(lang => ({
    term: translations[term][lang],
    lang: lang,
    api: apis[getRegionForLang(lang)]
  }));

  return Promise.all(queries.map(q => fetchFromAPI(q)));
}
```

### API Key Setup

```bash
# Europeana (Free)
export EUROPEANA_API_KEY=your_key_here

# Rijksmuseum (Free)
export RIJKSMUSEUM_API_KEY=your_key_here

# Harvard (from main report - Free)
export HARVARD_API_KEY=your_key_here
```

---

## 📥 Download Instructions

### 1. FANCY Dataset (302K images) - PRIORITY

**Steps:**
1. Visit: https://drive.google.com/drive/folders/1abIiasmgCSdvNpEDJM9-2iaURlTQh8PY
2. Download entire folder (may require Google Drive desktop app for large download)
3. Extract to: `handbag_data/international_collections/france/FANCY/`
4. Size: ~50GB estimated

**Includes:**
- All French brands: Hermès, Chanel, LV, Dior, etc.
- All Italian brands: Gucci, Prada, Fendi, Versace, etc.
- 146 attributes per image
- 25 style groups

**Alternative - ADM Dataset (smaller):**
- 25,470 images: https://drive.google.com/drive/folders/1rvd5D0vv2MrMEaZZpmeJ2ny6Tn44DyA-

---

### 2. Europeana Fashion API

**Setup:**
```bash
# 1. Register
# Visit: https://pro.europeana.eu/page/get-api
# Fill out simple form

# 2. Get API key (instant)

# 3. Set environment variable
export EUROPEANA_API_KEY=your_key

# 4. Run fetch script
node scripts/fetch_international_museum_data.js
```

**Expected Data:**
- Palais Galliera items
- French regional museums
- Italian fashion museums
- Dutch collections

---

### 3. Rijksmuseum API

**Setup:**
```bash
# 1. Register
# Visit: https://data.rijksmuseum.nl/

# 2. Request API key (instant approval)

# 3. Set environment variable
export RIJKSMUSEUM_API_KEY=your_key

# 4. Run fetch script
node scripts/fetch_international_museum_data.js
```

---

### 4. KCI Kyoto Manual Download

**Online Access (300 items):**
1. Visit: https://www.kci.or.jp/en/archives/digital_archives/
2. Browse chronologically
3. Save images + metadata manually
4. Or use web scraping (respect robots.txt)

**Full Access (13,000 items):**
- Requires on-site visit to Kyoto, Japan
- Study Room by reservation
- For serious research projects

---

### 5. Gallica BNF Manual Search

**Web Interface:**
1. Visit: https://gallica.bnf.fr/
2. Search: "sac Hermès" or "maroquinerie"
3. Filter: Images, fashion magazines
4. Download items individually

**Bulk Access:**
- Use Gallica API for automated access
- IIIF image endpoints available
- Requires understanding of French cataloging

---

## 🎯 Recommended Workflow

### Phase 1: Immediate (No API Keys)
1. ✅ Download FANCY dataset from Google Drive (302K images)
   - French brands: Complete
   - Italian brands: Complete
2. ✅ Browse KCI Kyoto online (300 items)
3. ✅ Manual Gallica searches for historical French data

**Expected Data:** 300K+ images, primarily runway/editorial

---

### Phase 2: API Access (Free registration)
1. Register for Europeana API key (5 minutes)
2. Register for Rijksmuseum API key (5 minutes)
3. Run automated fetch scripts
4. Process multilingual metadata

**Expected Data:** 1,000-10,000+ museum items

---

### Phase 3: Advanced (Optional)
1. Paris Musées GraphQL queries
2. Gallica API integration
3. KCI on-site research visit
4. Purchase commercial datasets (Gucci, LV)

---

## 📈 Brand Coverage Summary

### French Luxury Brands
| Brand | FANCY | Europeana | Gallica | KCI | Total Sources |
|-------|-------|-----------|---------|-----|---------------|
| Hermès | ✅ | ✅ | ✅ | ✅ | 4 |
| Chanel | ✅ | ✅ | ✅ | ✅ | 4 |
| Louis Vuitton | ✅ | ✅ | ✅ | Limited | 3+ |
| Dior | ✅ | ✅ | ✅ | Limited | 3+ |
| Celine | ✅ | ✅ | Limited | Limited | 2+ |
| YSL | ✅ | ✅ | ✅ | Limited | 3+ |
| Givenchy | ✅ | ✅ | Limited | Limited | 2+ |
| Balenciaga | ✅ | ✅ | Limited | ✅ | 3+ |

### Italian Luxury Brands
| Brand | FANCY | Europeana | Other | Total Sources |
|-------|-------|-----------|-------|---------------|
| Gucci | ✅ | ✅ | Commercial | 3 |
| Prada | ✅ | ✅ | Limited | 2+ |
| Fendi | ✅ | ✅ | Limited | 2+ |
| Versace | ✅ | Limited | Limited | 1+ |
| Bottega Veneta | ✅ | Limited | Limited | 1+ |
| Ferragamo | ✅ | ✅ | Limited | 2+ |
| Valentino | ✅ | ✅ | Limited | 2+ |

**Legend:**
- ✅ = Confirmed significant coverage
- Limited = Some items available
- Commercial = Available for purchase

---

## 🎁 Data Quality Assessment

### FANCY Dataset
- **Quality:** ⭐⭐⭐⭐⭐ Excellent
- **Coverage:** Comprehensive runway/editorial
- **Timeframe:** Contemporary (2010-2019)
- **Best For:** Modern luxury brand analysis, ML training

### Europeana Fashion
- **Quality:** ⭐⭐⭐⭐ Very Good
- **Coverage:** Historical + contemporary museum pieces
- **Timeframe:** 18th century to present
- **Best For:** Historical research, provenance

### Gallica BNF
- **Quality:** ⭐⭐⭐⭐ Very Good (archival)
- **Coverage:** Historical French fashion documents
- **Timeframe:** 18th-20th century strong
- **Best For:** Historical brand research, vintage

### KCI Kyoto
- **Quality:** ⭐⭐⭐⭐⭐ Museum-grade
- **Coverage:** Haute couture focused
- **Timeframe:** 17th century to present
- **Best For:** Designer pieces, Japanese perspective

### Rijksmuseum
- **Quality:** ⭐⭐⭐⭐ Very Good
- **Coverage:** Dutch + international
- **Timeframe:** Multiple centuries
- **Best For:** Historical European fashion

---

## 💾 Storage Requirements

| Source | Size Estimate | Priority |
|--------|---------------|----------|
| FANCY Dataset | ~50GB | HIGH |
| ADM Dataset | ~10GB | MEDIUM |
| Europeana API Results | ~500MB-2GB | HIGH |
| Rijksmuseum API Results | ~500MB-1GB | MEDIUM |
| KCI Manual Downloads | ~100MB | LOW |
| Gallica Items | Varies | LOW |

**Total Estimated:** 60-70GB for complete collection

---

## 🚀 Next Steps

### Immediate Actions:
1. **Download FANCY dataset** from Google Drive
   - Priority: HIGH
   - Impact: 302K images of French/Italian brands

2. **Register for API keys:**
   - Europeana (5 min) → French museum data
   - Rijksmuseum (5 min) → Dutch museum data

3. **Run automated fetches:**
   ```bash
   export EUROPEANA_API_KEY=your_key
   export RIJKSMUSEUM_API_KEY=your_key
   node scripts/fetch_international_museum_data.js
   ```

### This Week:
4. Browse KCI Kyoto online collection (300 items)
5. Manual Gallica searches for historical Hermès, Chanel
6. Process and normalize multilingual metadata

### This Month:
7. Integrate translation system
8. Build unified search across all sources
9. Create comprehensive French brand database

---

## 📞 API Support Contacts

**Europeana:**
- Email: api@europeana.eu
- Documentation: https://pro.europeana.eu/page/apis

**Rijksmuseum:**
- Email: api@rijksmuseum.nl
- Documentation: https://data.rijksmuseum.nl/

**Paris Musées:**
- Contact: Via website form
- API: https://apicollections.parismusees.paris.fr/

**KCI Kyoto:**
- Email: Via contact form on website
- Study Room: Reservation required

---

## ⚖️ Usage Rights Summary

### FANCY Dataset
- ✅ Non-commercial research only
- ❌ Commercial use prohibited
- ✅ Academic papers allowed with citation

### Europeana
- Varies by item
- Many CC0 or CC-BY
- Check individual item licenses

### Rijksmuseum
- ✅ CC0 for metadata
- Varies for images
- Generally permissive

### Gallica BNF
- Public domain for old works
- Check individual items
- Generally research-friendly

### KCI Kyoto
- Study/research use
- High-resolution requires permission
- Credit required

---

## 📊 Expected Final Dataset

After completing all downloads and API fetches:

**Total Estimated Items:**
- Images: 305,000+ (mostly FANCY)
- Museum Records: 5,000-15,000+
- Historical Documents: 500-2,000+
- **Grand Total: 310,000-320,000+ items**

**French Brand Coverage:**
- Hermès: 1,000s of items
- Chanel: 1,000s of items
- Louis Vuitton: 1,000s of items
- Other French luxury: 1,000s of items

**Italian Brand Coverage:**
- Gucci: 1,000s of items
- Prada: 1,000s of items
- Fendi, Versace, etc.: 100s-1,000s each

---

**Document Created:** 2025-11-11
**Status:** Ready for implementation
**Priority:** Download FANCY dataset immediately for French/Italian brands