Persons in Context, or PiCo, is a knowledge model for describing person observations and (derived from them) uniquely identified persons. The knowledge model uses a number of existing ontologies:

• Schema.org is used as much as possible for classes and properties
• BIO is used for some life events, such as marriage or divorce
• PNV is optionally used for extended person names
• PROV-O is used for information about the creation of person reconstructions

A limited number of proprietary properties are added. Inspired by ROAR, PiCo is based on mentions of persons in sources. From one or more person mentions a unique identifiable person can be reconstructed.

Introduction  

PiCo was developed and is managed by the CBG|Centre for family history for describing its own collections and as a new standard for submitting data to the WieWasWie platform.

The primary purpose of the knowledge model is to be able to search for ancestors. Person Reconstructions are uniquely identifiable within PiCo and linked to one or more Person Observation records. The development of PiCo was desired for the following reasons:

• New techniques allow us to enforce higher quality data, enabling new applications, such as the automated creation of uniquely identifiable persons;
• The CBG would like to make use of the possibilities of Linked Data as it has been embraced by cultural heritage institution in the Netherlands and abroad;
• Linked Data makes data sets easily extensible and applicable in other contexts;
• For a number of new sources and applications, the application of A2A at archival institutions turned out to be possible only with a (too) broad interpretation of the standard.

Three Concepts

To exchange data about persons in sources, PiCo distinguishes between three different concepts: a Person Observation, a Person Reconstruction and a Source.

Person Observation

A Person Observation is the representation of how a person is recorded in an original source. For example, a marriage certificate sometimes lists as many as ten people: in addition to the bride and groom, the couple's parents and witnesses.

Person Observations include data as much as possible as it appears in the source, that is, with the age, occupation, and spelling of the person's name as recorded in the deed. The model allows for choices when registering a Person Observation. For example, the creator of the Person Observation may include the date according to the ISO standard (for example: '1802-07-29') even though it is not present on the deed in this form (for example: '28 thermidor X').

Person Reconstruction

One or more Person Observations can be combined to create a uniquely identifiable person. We call this a Person Reconstruction. It can be constructed either automatically or by (more classical) genealogical research. When making a Person Reconstruction, the researcher may have made data more precise, for example by relating place names to standardized place name lists (such as Geonames). Each Person Reconstruction should include references to the Person records on which it is based.

Source

Each Person Observation is found in a Source. In addition to formal deeds and registrations, these can be books or personal documents such as letters, diaries or e-mails. Of a Source is always shown in which heritage institution the Source is kept and hopefully a link to the web address through which the Source is made available online. In PiCo, the resource type is indicated by the appropriate schema type, often schema:ArchiveComponent, but sometimes also schema:Book, schema:Photograph or even schema:Painting.

Support

Are you planning to use PiCo for publishing your historical person data? That's wonderful! It helps in connecting collections from different institution. That makes them more valuable and more eaily accessible to users.

Some tips for your implementation:

Be persistent

Persons in Context is a Linked Data standard. This means that every Person Observation (or every Person Reconstruction, if you have any) should have an identifier that is in fact a URL. Keep in mind that every user of your data, whether it is a private genealogists, a professional researcher or an aggregation platform, should be able to rely on the consistency of those URL's. Now and in the future. Otherwise, the results of their hard work will lose the connection with the sources it relies on. Your sources.

To learn more about persistent identifiers, the following links are very useful. They are all provided by the Dutch Digital Heritage Network

• Three short video's containing a basic explainer about why and how to use Persistent Identifiers: Sustained access with Persistent Identifiers
• A brief explainer about the why and how of persistent identifers (in Dutch): Persistent Identifiers
• Some information about maintaining persistent identifiers: Werk je al met persistent identifiers? Met deze tips maak je het beheer ervan nog makkelijker

For the technical implementation of persistent identifiers, you'll probably want to refer to the supplier of your collection management system. All common collection management systems in the Netherlands (Atlantis, TMS, Mais Flexis, Adlib/Axiell Collections en Memorix) support them.

We strongly advise you to read the Persistent Identifier Guide. It helps you to choose the right PID System for your organisation. And the FAQ and the Best Practice give a good impression of the impact of using PID's in your organisation. Spoiler: it takes some initial effort to get things going, but once you're done there is not much to it. And it will make your collection much more usable!

Use IIIF for linking to scans

What goes for records, goes for scans as well: the links you provide to the scans of your sources should be persistent. The ideal situation would be where the link to a scan is a persistent URL pointing to a IIIF manifest.

Serving your images in IIIF format is a technical effort. The provider of your collection management system will most likely be able to help you out. Some general information about IIIF:

• NDE over IIIF (Dutch)
• Official IIIF website (English)

Documentation

Please refer to the official PiCo documentation page (link below). It contains a description of all classes and properties used within the model, schematic overviews and examples.

Reporting issues

If you found an issue in the PiCo model or somewhere in the documentation, you can create an issue on he PiCo GitHub page . Or email us at personsincontext@cbg.nl

Validating your PiCo data

A SHACL file is available for validating your PiCO graph data. Keep in mind that this is a very basic validation at this moment. More validation rules will be added over time. There is a download link below. A README file for the usage of the SHACL file can be found on the PiCo Github page.

Convert A2A or GEDCOM to PiCo

Conversion scripts are available on GitHub for converting A2A or GEDCOM data to PiCo:

• A2A to PiCo
• GEDCOM to PiCo

Further reading

Articles published online about Persons in Context and its applications:

• Persons in Context. A Model to Represent Observations and Reconstructions of Historical Persons in Linked Data (English)
• Nieuwe standaard CBG voorziet in behoefte (Dutch)
• PiCo in Practice (English)