Working Draft, 03 January 2023

This version: https://purl.org/episa/archonto

Latest version: https://purl.org/episa/archonto

Previous version: https://purl.org/episa/archonto/0.0.6

Abstract

This specification describes ArchOnto, a linked open data model. ArchOnto adopts a semantic web approach and encompasses the CIDOC Conceptual Reference Model and additional ontologies, envisioning interoperability with datasets curated by multiple communities of practice. Its goal is to serve as the foundation for the development of a knowledge graph prototype and new applications for the archives.

Status of this document

This document is in progress. This may undergo changes considering that the present ontology is in constant development.

Currently, the ArchOnto can already represent the description of archival records, identifying people, places, dates, and related events. In the future, the ontology will represent the digital representations associated with each document.

The ArchOnto specification is produced as part of the Entity and Property Inference for Semantic Archives (EPISA) project at INESC TEC to provide authoritative documentation of the contents, status and purpose of the linked data model known as ‘ArchOnto’.

As a complement to the ArchOnto specification, a related document - Individuals - identifies individuals common to various archival descriptions. These can be imported when needed to make the description of archival objects.

Table of Contents

• Glossary - In progress
• Introduction - In progress
• Architecture - In progress
• Classes and Properties - In progress
  • Archonto - In progess
  • DataObject - In progress
  • N-ary - In progress
  • CIDOC CRM - In progress
  • ISAD Ontology - In progress
  • Link2DataObject - In progress

Glossary

ArchOnto

• Ontology created in the context of archives as an extension to CIDOC-CRM.
• Uses the prefixes:
  • ARE – Archival Entity, eg ARE1 Level of Description.
  • ARP – Archival Property, eg APR12 has level of description.

CIDOC-CRM – CIDOC Conceptual Reference Model

• Reference ontology, designed to facilitate the integration, mediation and exchange of information in the field of cultural heritage.
• Provides the semantic definitions and rules needed to transform fonts.
• localized and disparate information into a coherent global resource.
• CRM intends to serve as a reference in terms of concepts and, therefore, of semantics to database schemas and organizational structures of
• documents in various information systems.
• Use the prefixes:
  • E – Entity, eg: E1 CRM Entity;
  • P – Property, eg: P1 is identified by

DataObject

• Ontology created in order to validate literal values used in properties of the newly created data model.
• It has classes and Data Properties in its base that are used to validate data from simple types in the ontology, ensuring class-based validation.

ISAD (G) - International Standard for Archival Description

• Establishes general guidelines for the preparation of archival descriptions that can be applied to any type of document, regardless of its dimension, shape or support.

ISAD Ontology

• Ontology created to represent all fields of the ISAD (G) standard;
• It has Data Properties in its base that allow the incorporation of all fields ISAD(G). With this ontology, the information of archival descriptions is kept existing files created over the years by the various archives at the national level.
• The ontology allows, whenever necessary, the validation of the contents that were atomized, verifying if they are in accordance with the registration that already existed in the ISAD description (G).
• All information associated with a unit of description is considered as a physical object; in CIDOC-CRM there are attributes associated with the physical object and others with the conceptual object. In this ontology, it is assumed that the physical object incorporates the conceptual object.
• With this ontology it will be possible to maintain the interoperability of all data represented in ISAD (G).

Link2DataObject

• Ontology created to link the CIDOC-CRM ontology with the ontology DataObject.
• Imports both ontologies and has an Object Property that allows to link the two imported ontologies – hasValue.

N-ary

• Ontology created to represent arity tuples greater than 2.
• With this ontology it is possible to represent all instances where it is necessary to represent relationships that link more than two individuals.
• Ontology created taking into account the suggestion given by CIDOC-CRM to represent non-binary relationships.

1. Introduction

ArchOnto is the name of the new model supporting the representation of assets at DGLAB. Its goal is to serve as the foundation for developing a knowledge graph prototype and new applications for the archives.

ArchOnto is a modular data model that integrates five ontologies, dealing with complementary aspects of the archival domain. The five ontologies are CIDOC CRM (base ontology), N-ary, DataObject, ISAD Ontology, and Link2DataObject. ArchOnto also has its classes and properties.

In the figure below there is an application example of ArchOnto ontology.

2. Architecture

ArchOnto is a modular data model that integrates five ontologies, dealing with complementary aspects of the archival domain, as illustrated in figure 1. The five ontologies are:

• CIDOC-CRM (base ontology): the core of ArchOnto, providing the concepts and properties to capture archival records’ essential features, e.g., event, date, location, person, group.

• N-ary: provides a systematic way to represent non-binary associations, i.e., those that connect more than two individuals. Representing a person’s role in a given event is an example of this kind of association because it involves relating an event, a person, and a role.

• DataObject: auxiliary ontology to deal with literal values and their validation in ArchOnto. Comprises classes and data properties for the simple types in the ontology. In this ontology, the names of classes are preceded by DOE (from DataObject Entity), and the names of properties are preceded by DOP (from DataObject Property).

• ISAD Ontology: matches the ISAD(G) standard elements that can be directly obtained from existing records. The information is represented with Data Properties and is not atomized. The ontology can be regarded as an intermediate representation for legacy records, making it possible, whenever necessary, to check the contents of a given element against the information extracted thereof. This ontology is only used for the existing records in DigitArq and not for newly created records based on ArchOnto.

• Link2DataObject: makes the connection between the CIDOC-CRM and the DataObject ontologies with a single property to make this connection — L2DO hasValue.

In addition to these five ontologies, ArchOnto has its own classes and properties, created as extensions to CIDOC-CRM. Classes are used to specialize concepts already present in CIDOC-CRM but which needed to be adapted to include controlled vocabularies that exist in the archives. CIDOC-CRM. In addition to these classes, the ontology also includes three properties which provides a control structure for the admissible nesting of units of description. Names for classes are preceded by ARE (from ARchival Entity), and those for properties are preceded by ARP (from ARchival Property).

Figure 1. ArchOnto architecture

3. Classes and Properties

Full detail regarding each class and property can be found in the following subsections: ArchOnto, DataObject, N-ary, CIDOC-CRM, ISAD Ontology and Link2DataObject.

Class hierarchy within ArchOnto’s ontologies

Property hierarchy within ArchOnto’s ontologies

ArchOnto

ArchOnto is a modular data model that integrates five ontologies, dealing with complementary aspects of the archival domain. The five ontologies are: - CIDOC-CRM (base ontology); - N-ary; - DataObject; - ISAD Ontology; - Link2DataObject.

In addition to these five ontologies, ArchOnto has its own classes and properties, created as extensions to CIDOC CRM. Classes are used to specialize concepts already present in CIDOC CRM but which needed to be adapted to include controlled vocabularies that exist in the archives. In addition to these classes, the ontology also includes three properties which provides a control structure for the admissible nesting of units of description. Names for classes are preceded by ARE (from ARchival Entity), and those for properties are preceded by ARP (from ARchival Property).

Classes

• ARE1 Level of Description
• ARE2 Formal Title
• ARE3 Supplied Title
• ARE4 Extension
• ARE5 Identifier Type
• ARE6 Date Type
• ARE7 Name Type
• ARE8 Role Type
• ARE9 Date Certainty
• ARE11 Documentary Typology
• ARE12 Organisation
• ARE13 Subject Type
• ARE14 Place Type
• ARE15 Acquisition Type
• ARE16 Event Type

Properties

• ARP8 upper level
• ARP9 lower level
• ARP12 has level of description