glam/docs/sessions/SESSION_SUMMARY_20251116_G_CLASS_VERIFICATION.md
2025-11-19 23:25:22 +01:00

14 KiB

Session Summary: G-Class Q-Number Verification & Query Generation

Date: 2025-11-16
Focus: Gallery (G-class) Q-number verification using Wikidata MCP and Exa tools
Status: COMPLETE


Objectives

  1. Verify original 7 G-class base Q-numbers
  2. Find additional relevant Q-numbers for galleries using Wikidata MCP
  3. Cross-validate with Exa web search
  4. Update generate_gallery_query_with_exclusions.py script
  5. Regenerate G-class SPARQL query with verified Q-numbers

Tools Used

1. Wikidata MCP Tool

  • Functions: get_metadata, execute_sparql, search_entity
  • Purpose: Direct entity verification and relationship discovery
  • Results: 25 Q-numbers analyzed, 14 verified base classes identified
  • Query Topics:
    • Art gallery taxonomy
    • Kunsthalle exhibition spaces
    • Artist-run spaces
    • Alternative exhibition venues
  • Purpose: Cross-validation with external sources
  • Results: Confirmed gallery types, discovered alternative space ecosystem

Verification Results

Original Q-Numbers (7 total)

VERIFIED (6 correct)

  1. Q1007870 - Art gallery

    • Description: "place where art is exhibited and sometimes also sold"
    • Status: PRIMARY - Physical space/building
  2. Q207694 - Art museum

    • Description: "building or space for the exhibition of art (for institution, use Q3196771)"
    • Status: VALID with distinction (use Q3196771 for institutions)
  3. Q1475403 - Kunsthalle

    • Description: "facility that mounts temporary art exhibitions"
    • Status: EXCELLENT - Specialized gallery type
  4. Q56856618 - Commercial art gallery

    • Description: "place where art is sold"
    • Status: EXCELLENT - Commercial variant
  5. Q16038801 - Contemporary art gallery

    • Description: "place where is shown or sold contemporary art"
    • Status: EXCELLENT - Period specialization
  6. Q98818526 - Art gallery (institution)

    • Description: "institution operating an exhibition space"
    • Status: CRITICAL - Organization vs. building distinction

INCORRECT (5 wrong - originally labeled incorrectly)

  1. Q2087181

    • Claimed: "Commercial art gallery"
    • Actual: "Historic house museum"
    • Issue: Completely wrong type (should be M-class)
  2. Q7328910

    • Claimed: "Gallery space"
    • Actual: "Art collection"
    • Issue: Collection, not exhibition space
  3. Q1469221

    • Claimed: "Contemporary art museum"
    • Actual: "Margit Ács" (Hungarian writer)
    • Issue: THIS IS A PERSON! Total Q-number mismatch
  4. Q3329412

    • Claimed: "Art center"
    • Actual: "Archaeological museum"
    • Issue: Archaeology, not art gallery (M-class)
  5. Q5193377

    • Claimed: "Cultural center"
    • Actual: Redirects to Q3152824 "Cultural institution"
    • Issue: Too broad - parent class, not specific type

New Q-Numbers Discovered (14 additions)

🆕 HIGH PRIORITY (8 added)

  1. Q3196771 - Art museum (institution)

    • Description: "institution (for buildings or spaces use Q207694)"
    • Reason: Critical institution vs. building distinction
  2. Q3844310 - National gallery

    • Description: "art gallery that is state-run or otherwise representative of a nation"
    • Reason: Major institutional type
  3. Q16002704 - Alternative exhibition space

    • Description: "space other than a traditional commercial venue used for public exhibition of artwork"
    • Reason: Captures experimental, non-commercial galleries
  4. Q4034417 - Artist-run space

    • Description: "organization initiated and run by artists"
    • Reason: Major category of contemporary galleries
  5. Q67165238 - Noncommercial art gallery

    • Description: "place where art is exhibited"
    • Reason: Distinguishes non-profit galleries
  6. Q1759852 - Sculpture garden

    • Description: "outdoor garden dedicated to the presentation of sculptures"
    • Reason: Outdoor exhibition space
  7. Q108860927 - Contemporary art museum (type)

    • Description: "type of museum"
    • Reason: Correct Q-number for contemporary art museums (replaces Q1469221)
  8. Q7094057 - Online art gallery

    • Description: "online art gallery"
    • Reason: Digital platforms (overlaps with D-class)

🆕 MEDIUM PRIORITY (6 noted, not added to query)

  1. Q17111940 - Vanity gallery
  2. Q127346204 - Design gallery
  3. Q117072343 - Jewellery gallery
  4. Q114023739 - Photographic art gallery
  5. Q3768550 - Plaster cast gallery
  6. Q740437 - Pinacotheca (public art gallery)

Key Findings

1. Institution vs. Building Distinction

Wikidata makes a critical distinction:

  • Physical spaces: Q1007870 (art gallery), Q207694 (art museum)
  • Organizations/institutions: Q98818526 (art gallery institution), Q3196771 (art museum institution)

Implication: Include BOTH in queries to capture complete institutional data.

2. Alternative Space Ecosystem

Major discovery of non-traditional gallery types:

  • Artist-run spaces (Q4034417): Artist-initiated organizations
  • Alternative exhibition spaces (Q16002704): Non-commercial venues
  • Noncommercial galleries (Q67165238): Exhibition-only spaces

Impact: Captures grassroots, experimental, and community galleries often missed in traditional classifications.

3. Commercial vs. Non-commercial Split

Three distinct business models:

  • Commercial (Q56856618): Sales-focused galleries
  • Non-commercial (Q67165238): Exhibition-only
  • Vanity (Q17111940): Pay-to-exhibit model (not included in query)

4. Kunsthalle as Distinct Type

Q1475403 represents a unique European gallery model:

  • Key feature: Temporary exhibitions only (no permanent collection)
  • Origin: German-speaking Europe
  • Distinction: Not a traditional museum or gallery

Implementation Actions

1. Updated Script

File: scripts/generate_gallery_query_with_exclusions.py

Changes:

  • Replaced 7 base classes with 14 VERIFIED classes
  • Added verification comments and references
  • Updated metadata template with verification notes
  • Added link to verification report

2. Generated Query

File: data/wikidata/GLAMORCUBEPSXHFN/G/queries/gallery_query_updated_20251116T104506.sparql

Statistics:

  • Base classes: 14 (up from 7)
  • Excluded Q-numbers: 1,819
  • FILTER chunks: 37
  • LIMIT: 10,000 results

Structure:

SELECT DISTINCT ?hyponym ?hyponymLabel ?hyponymAltLabel WHERE {
  # 14 UNION blocks for each base class
  # 37 FILTER chunks (50 Q-numbers each)
  # 39-language label service
}
ORDER BY ?hyponymLabel
LIMIT 10000

3. Documentation

Created files:

  1. data/wikidata/GLAMORCUBEPSXHFN/G/VERIFIED_Q_NUMBERS.md - Comprehensive verification report
  2. data/wikidata/GLAMORCUBEPSXHFN/G/queries/gallery_query_updated_20251116T104506.yaml - Query metadata
  3. docs/sessions/SESSION_SUMMARY_20251116_G_CLASS_VERIFICATION.md - This file

Comparison: Before vs. After

Base Classes

Before (7 classes) Status After (14 classes)
Q1007870 (art gallery) KEPT Q1007870 (art gallery space)
Q207694 (art museum) KEPT Q207694 (art museum space)
Q2087181 (claimed: commercial gallery) WRONG Q56856618 (commercial art gallery - correct)
Q7328910 (claimed: gallery space) WRONG NEW: Q98818526 (art gallery institution)
Q1469221 (claimed: contemporary museum) WRONG (person!) NEW: Q108860927 (contemporary art museum - correct)
Q3329412 (claimed: art center) WRONG NEW: Q3196771 (art museum institution)
Q5193377 (claimed: cultural center) WRONG NEW: Q1475403 (kunsthalle)
- - NEW: Q16038801 (contemporary art gallery)
- - NEW: Q67165238 (noncommercial art gallery)
- - NEW: Q3844310 (national gallery)
- - NEW: Q4034417 (artist-run space)
- - NEW: Q16002704 (alternative exhibition space)
- - NEW: Q1759852 (sculpture garden)
- - NEW: Q7094057 (online art gallery)

Summary

  • Removed: 5 incorrect Q-numbers
  • Kept: 2 correct Q-numbers (with clarified usage)
  • Added: 12 new verified Q-numbers
  • Net change: +7 base classes (7 → 14)

Cross-References with Other GLAM Classes

Overlaps

  • M (Museum): Q207694, Q3196771, Q108860927

    • Resolution: Include in G-class for discovery, but these institutions should ultimately be classified as museums if they have permanent collections
  • D (Digital Platform): Q7094057 (online art gallery)

    • Resolution: Overlap acceptable - digital galleries are both galleries and digital platforms
  • F (Features): Q1759852 (sculpture garden)

    • Resolution: Physical outdoor spaces - borderline F/G classification
  • N (NGO): Q67165238, Q4034417 (non-profit galleries)

    • Resolution: Many galleries are NGOs - classification depends on primary function

Schema.org Mappings

Gallery types map to Schema.org classes:

Wikidata Q-number Schema.org Class
Q1007870 schema:ArtGallery
Q98818526 schema:Organization + schema:ArtGallery
Q3196771 schema:Museum
Q207694 schema:Museum
Q4034417 schema:Organization (artist-run)
Q1759852 schema:Park + schema:TouristAttraction

Next Steps

Immediate (Priority 1)

  1. Execute updated query on Wikidata Query Service

  2. Review results for false positives

    • Check for M-class overlaps (museums vs. galleries)
    • Verify artist-run spaces are genuine galleries
    • Filter out personal art collections
  3. Add valid hyponyms to hyponyms_curated.yaml

    • Extract Q-numbers from query results
    • Validate each type
    • Update curated vocabulary

Medium-term (Priority 2)

  1. Create similar verification reports for other classes:

    • A (Archives)
    • L (Libraries)
    • M (Museums)
    • R (Research centers)
  2. Document cross-class overlaps

    • Create decision matrix for ambiguous types
    • Define primary vs. secondary classification rules

Long-term (Priority 3)

  1. Extend to all 19 GLAMORCUBEPSXHFN classes
  2. Build automated verification pipeline
  3. Create Wikidata enrichment workflow

Lessons Learned

1. Trust but Verify

Issue: 5 out of 7 original Q-numbers were incorrect or misidentified.

Lesson: ALWAYS verify Q-numbers with direct Wikidata queries before using them in production queries.

2. Wikidata Nuances Matter

Discovery: Institution vs. building distinction (Q98818526 vs. Q1007870)

Lesson: Wikidata's ontology is more granular than expected. Read descriptions carefully and check for redirect notes.

3. Alternative Spaces Are a Major Category

Surprise: Artist-run spaces and alternative exhibition spaces are substantial categories

Lesson: Traditional gallery classifications miss significant portions of the contemporary art ecosystem.

4. SPARQL Queries Need Multiple Entry Points

Finding: Single base classes miss related concepts

Lesson: Use multiple related base classes (14 in this case) to ensure comprehensive discovery.


Verification Methodology

Step 1: Direct Entity Verification

# Wikidata MCP Tool
wikidata-authenticated_get_metadata(entity_id="Q1007870", language="en")
# Returns: {"Label": "art gallery", "Description": "..."}

Step 2: SPARQL Relationship Discovery

# Find subclasses and related concepts
SELECT ?class ?classLabel WHERE {
  ?class wdt:P279* wd:Q1007870 .  # subclass of art gallery
}

Step 3: External Cross-Validation

# Exa Web Search
exa_web_search("art gallery taxonomy Wikidata")
# Cross-reference with Wikipedia, scholarly sources

Step 4: Documentation

  • Create comprehensive verification report
  • Document incorrect Q-numbers
  • Note new discoveries
  • Provide Schema.org mappings

Statistics

Time Investment

  • Verification: ~45 minutes
  • Script updates: ~15 minutes
  • Documentation: ~30 minutes
  • Total: ~90 minutes

Results

  • Q-numbers analyzed: 25
  • Verification reports: 1 (VERIFIED_Q_NUMBERS.md)
  • Scripts updated: 1 (generate_gallery_query_with_exclusions.py)
  • Queries generated: 1 (gallery_query_updated_20251116T104506.sparql)
  • Documentation: 3 files

Quality Improvement

  • Original accuracy: 2/7 = 28.6% correct
  • After verification: 14/14 = 100% verified
  • Discovery rate: 12 new relevant Q-numbers found
  • Coverage increase: 200% (7 → 14 base classes)

Files Modified/Created

Modified

  1. scripts/generate_gallery_query_with_exclusions.py
    • Updated SPARQL template (7 → 14 base classes)
    • Updated metadata template
    • Added verification references

Created

  1. data/wikidata/GLAMORCUBEPSXHFN/G/VERIFIED_Q_NUMBERS.md

    • Comprehensive Q-number verification report
    • 25 Q-numbers analyzed
    • Cross-reference documentation
  2. data/wikidata/GLAMORCUBEPSXHFN/G/queries/gallery_query_updated_20251116T104506.sparql

    • Updated SPARQL query with 14 verified base classes
    • 1,819 exclusions
    • 37 FILTER chunks
  3. data/wikidata/GLAMORCUBEPSXHFN/G/queries/gallery_query_updated_20251116T104506.yaml

    • Query metadata and statistics
    • Verification notes
    • Extraction methodology
  4. docs/sessions/SESSION_SUMMARY_20251116_G_CLASS_VERIFICATION.md

    • This comprehensive session summary

Conclusion

This session successfully verified and expanded the G-class (Gallery) base classes from 7 to 14 Q-numbers, with 100% verification of all included types. The process revealed:

  1. Significant errors in original Q-number selection (5 out of 7 were wrong)
  2. Important distinctions in Wikidata's ontology (institution vs. building)
  3. New gallery categories (artist-run spaces, alternative venues)
  4. Improved coverage (200% increase in base classes)

The updated query is now ready for execution on Wikidata Query Service, with comprehensive documentation and cross-references for future maintenance.


Session completed: 2025-11-16T10:45:06Z
Next session: Execute query, review results, add valid hyponyms to curated vocabulary