Open Data & Ontologies

A key aspect of the nanohistory method is modelling and reuse of the data created by users. The platform's data is designed to be both human and machine readable.

• Open Data & Ontologies
• Identifier Types
• Import Templates

An 'ontology' in the world of open data is a description of the relationship of data elements such as entities, attributes, fields, etc. within a specific context such as a data set, or involving other ontologies or vocabularies. As a prototype, Nanohistory's data model is still in flux as how its networked approach to historical representation relies on nested named graphs. Nested named graphs are not common in RDF, an present challenges in describing the model in OWL, the usual language for expressing ontologies. Further, Nanohistory's area of focus - events and actions - remains fluid. Almost all existing event models 'reify' events, or have other problems (mereology, for instance), that conflict with historiographical theories. This is the fundamental challenge of the platform, in many ways. The unique approach of Nanohistory means it serves to complement the richness offered by core ontologies such as CIDOC-CRM, but does so with a focus on processes and actions, rather than on representation of objects or things, as its first-principle. Building out an ontology for Nanohistory using methods such as RDF-Star does appear to provide a workable solution: the resulting ontology is massive. Due to these issues, Nanohistory is now exploring Labeled Property Graph approaches for describing its model.

The most recent expression of Nanohistory's ontology breaks the model into two separate ontologies: a CORE ontology where the fundamentals are expressed without specific verbs, and a child or secondary ontology containing a list of verbs that are currently in use as edges in the nested named graphs. Neither is production grade-ready. However, this version moves the needle down the road towards a clearer solution. The CORE Ontology is accessible at http://purl.org/nanohistory/ns, while the Verb Ontology at http://purl.org/nanohistory/verbs. The verb ontology consists mainly of properties for verbs; it is extensive and much larger than the CORE ontology. It can be visualized using WebOWL VOWL at https://service.tib.eu/webvowl/#iri=http://purl.org/nanohistory/verbs. A visualization of the current version 2.0 CORE ontology using WebOWL VOWL is below.

Here is an example of the kind of open data nanohistory can generate. This particular example contains data around Christopher Dare's questions to Anne Askew about whether she believed in the efficacy of 'private masses' for the dead. It could, however, contain data on anything. Note the nested events - Christopher Dare questions Anne, about a particular belief, and she answers that question. The exchange is only mentioned in one of the three secondary sources used in the case study, Megan Hickerson's article.

In addition, each entity within the platform can be downloaded as a stand-alone record, in a number of standards.