Project Status - DDMAL/linkedmusic-datalake GitHub Wiki

Completed Work

A visualization of the current complete LinkedMusic data lake ontology can be found here.

As of 20 August 2025, we have finished ingesting the following databases to the LinkedMusic data lake, a total of 383,809,033 RDF triples:

Note: As our workflow becomes more streamlined, these databases would benefit from additional passes of the full ingestion pipeline

• MusicBrainz - an open music encyclopedia that provides extensive music metadata and serves as a universal reference for music identification

  • 355,651,218 RDF triples
  • visualization
  • documentation

• DIAMM - an archive of digital images of European medieval manuscripts

  • 483,631 RDF triples
  • visualization
  • documentation

• The Session - a community website dedicated to Irish traditional music

  • 1,052,162 RDF triples
  • visualization
  • documentation

• The Global Jukebox - a website focusing on traditional folk, indigenous, and popular songs from around the world

  • 110,044 RDF triples
  • visualization
  • documentation

• Dig That Lick 1000 - a project the extracts and analyses solos from jazz performances

  • 13,756 RDF triples
  • visualization
  • documentation

• CantusDB - a repository of Latin chants found in medieval manuscripts and early printed books

  • 3,570,624 RDF triples
  • visualization
  • documentation

• RISM - an international collaborative database that catalogues historical musical sources

  • 22,927,598 RDF triples
  • visualization
  • documentation

Datasets in-progress

• Renaissance Liturgical Imprints: A Census

• SIMSSA Database - a discovery tool for symbolic music files (MEI, Kern, MusicXML, MIDI)

  • documentation

• Cantus Index - a catalogue of liturgical chant texts and melodies

  • documentation

• AcousticBrainz - a collection of acoustic information from music recordings between 2015 and 2022, providing insights into spectral data, genres, moods, keys, and scales.

  • documentation

Datasets to ingest in the near future

• Weimar Jazz Database
• CritiqueBrainz
• ListenBrainz
• Cover Art Archive
• Digital Analysis of Chant Transmission
• Printed Sacred Music Database
• ESEA (East-and-Southeast-Asian) & Chinese (Traditional) Music Instrument - located in the store

Additional tasks

1. Develop the NLQ2SPARQL tool (SESEMMI)

• NLQ2SPARQL Context contains the currently used context when testing LLM SPARQL query generation
• Refer to Sample LinkedMusic Queries for queries across our four benchmark challenges used in testing
• Refer to NLQ2SPARQL and NLQ2SPARQL Q&A (Unfinished) for previous work

2. Upload new entities and properties to Wikidata

• Refer to Wikidata Uploading (Feast Day Project) for our current approach to uploading Wikidata, starting with Feasts
• Refer to Wikidata: Things we should add for other categories of items that would be useful for our purposes if they were on Wikidata

  Note: Dataset-specific documentation may have a partial list of specific items missing from Wikidata

3. Develop the front end for LinkedMusic