Chiara Latronico, The Golden Agents project. Disambiguating Person Name Entities
Download as PPTX, PDF0 likes76 views
The Golden Agents Project aims to enhance understanding of the Dutch Golden Age's creative industries through the analysis of production and consumption data. It utilizes semantic web and multi-agent technology to process and enrich vast datasets, including century-old registries and notarial acts. Funding has been secured for a five-year duration, focusing on disambiguating person name entities and connecting disparate historical data sources.
Chiara Latronico, The Golden Agents project. Disambiguating Person Name Entities
- 1. The Golden Agents project Disambiguating Person Name Entities Chiara Latronico, International Archival Conference, Čačak, 10-10-2019
- 2. Agenda ● About the Golden Agents project ● Amsterdam City Archive registries ● Transforming and basic interpreting. Processing and enriching the datasets. ● Disambiguating between Persons: Lenticular Lenses II ● Network analysis
- 3. About the Golden Agents project
- 4. Partners NWO-Large Infrastructure Project – funding ca. $ 4 million budget € 6 million : 2017 – 2021 (5 years)
- 5. Golden Agents Infrastructure: Aim Understanding the dynamics in the creative industries of the Dutch Golden Age ● By analyzing interactions between various branches of the creative industries ● By analyzing interactions between producers and consumers of the creative industries
- 6. Problem 1 data about production of creative industries dispersed over separated databases
- 7. Problem 2 lack of digital data about the consumption of the creative industries of the Dutch Golden Age
- 8. Golden Agents: Infrastructure ● Combines Semantic Web and Multi Agents technology to analyze and to interact with existing and new datasets (in linked data) about the Dutch Golden Age ● Develops ontologies incorporating these dynamic interactions as storylines and events in linked data ● Uses a combination of hand-written text recognition and crowdsourcing to disclose 2 million scans of notary acts
- 10. Amsterdam City Archives Registies ● Notarial Acts (Notarieel archief) ● Baptism Registries (Doopregisters) ● Burial Registries (Begraafregisters) ● Prenuptial Agreements (Ondertrouwregisters) ● Transport Acts (Kwijtscheldingen) ● Confession Books (Confessieboeken) ● Burghers Books (Poorterboeken) ● Fines on marriage and burying (Boetes op trouwen en begraven) https://www.amsterdam.nl/stadsarchief/organisatie/open-data/
- 11. Notarial Acts ● 3,5 km ● 731 notaries (1578-1915) ● 30.000 inventory n. ● 6.419.017 scans ● 783 volunteers ● Millions of deeds
- 13. Transforming and basic interpreting. Processing and enriching the datasets.
- 15. Basic Conversion XML to RDF ● Golden Agent Ontology ○ PREFIX saaOnt: <http://goldenagents.org/uva/SAA/ontology/> ■ saaOnt:identifier ■ saaOnt:mentionsBride ■ saaOnt:mentionsGroom ■ saaOnt:mentionsEarlierHusband ■ saaOnt:mentionsEarlierBride ○ Person names are stored as pnv:PersonName ■ No interpretation: PersonName = Name
- 16. Enrichment and Interpretation ● Event based data model ● PersonNames attached to Persons ● Relations between Persons are indicated ○ schema:spouse, bio:Marriage ● Implicit person mentions are added as Person ○ Child of ○ Widow of ● Georeferences are expressed with possible geometries ● Further information structuring if needed ○ Splitting StringLiterals into individual resources and types ● Duplicate records are removed.
- 17. Enrichment: Geo-Referencing link to historical and contemporary map
- 20. Disambiguating between Persons: Lenticular Lenses II
- 21. Disambiguating Person Names One of the several experiments ● Prenuptial Agreements ● Baptism Registries ● ECARTICO ECARTICO is a comprehensive collection of structured biographical data concerning painters, engravers, printers, book sellers, gold- and silversmiths and others involved in the ‘cultural industries’ of the Low Countries in the sixteenth and seventeenth centuries. (http://www.vondel.humanities.uva.nl/ecartico/)
- 35. Contextual Entity Disambiguation in Domains with Weak Identity Criteria: Disambiguating Golden Age Amsterdamers (Accepted)