RiC-O Converter: an example of practical application of the ICA Records in Contexts standard (ICA RiC)

English, IAW2020: Empowering knowlegde societies Leave a comment
Florence Clavaud

Florence Clavaud

ICA Records in Contexts (ICA RiC, abbreviated to RiC in this article) is the ICA’s new archival description standard, developed by EGAD (Expert Group on Archival Description).

  • RiC-O Illustration Conference
  • RiC-O conference Illustration

This standard, which is still in draft form, will ultimately combine and replace the four existing international archival description standards that were published between 1994 and 2008 (ISAD(G): ISAAR(CPF), ISDF and ISDIAH). It will provide an overarching framework adapted to changes in the profession and digital technologies. With this standard, better allowance can be made for the highly complex nature of archives and their history and for the numerous contextual layers inherent in archival collections.

First of all, RiC consists of an abstract conceptual model (RIC-CM). This model targets the profession and clarifies notions specific to the archival community. RiC-CM sets these notions out along similar lines to those used by other professional communities( 1). Using this model, the archival community will be able to communicate with other communities on clearly defined bases and to work with them to develop real interoperability for cultural data.

The second part of RiC (RiC-O) is an ontology, in other words a formal, technical representation of RiC-CM, consisting of a file setting out the vocabulary to be used and the rules to apply to digital archival metadata conforming to RiC-CM and having the form of RDF datasets publishable as linked or semantic web data. The relationship between RiC-O and RiC-CM is somewhat similar to that between XML/EAD and ICA ISAD(G)( 2).

Practical requirements for adopting RiC-O

Adopting RiC-O will call for a certain number of specific skills: familiarity with XML and RDF technologies (e.g. SPARQL query language), web ontology language (OWL), and at least basic knowledge of the technical solutions for storing, managing and publishing RDF datasets: RDF databases or triple stores. These are skillsets that do not always form part of archival studies curricula, even though archival professionals are increasingly familiar with at least some of these notions. To use RiC-O on a general or more limited basis within an institution, it will therefore be necessary to call on the services of experts with systems engineering profiles, specialising in web technologies and knowledge engineering.

This may all seem fraught with difficulties. But their will be significant benefits, from the degree of precision of the data obtained to the potential for establishing links with the data of other institutions, plus access to collections and the ability to make repeated use of the same data.

The example of French National Archives (ANF)

ANF archival lcollection users currently have to browse a number of unconnected websites and databases when conducting their research. The online catalogue is admittedly the main point of entry, but the many relationships between the objects described (mainly documents and their creators) are completely unpollable. Users cannot poll the system on a blanket basis for information about an individual (who?), a place (where?) or a subject (what?).

With RiC it is possible to represent an archival collection, all the documents that it includes, the people that produced it or that it concerns, the places involved, etc. in the form of a graph of interconnected entities. Each of the entities on the graph will then be a potential point of entry for the end user. In addition, some of these entities are known by institutions other than ANF and can therefore become access keys to the data contained in several different information systems. The result is descriptions in the form of multidimensional and dynamic graphs.

To begin with, we wanted to check this concept by demonstrating its validity via a small subset of our archival finding aids and authority controls. This produced the PIAAF prototype put on line in February 2018, and developed in partnership with BnF and SIAF. With this, we were able to demonstrate the feasibility of converting existing data into RDF according to RiC-O and to highlight the advantages of graphic archival metadata displays. Not only were the results obtained of high quality but we were also able to recognise the improvement this solution offers in terms of accuracy and search possibilities.

We have therefore decided to move to a much larger scale and make far more systematic use of this solution by acquiring the RiC-O Converter as ANF’s first software building block.

What changes will there be as a result of RiC-O Converter?

RiC-O Converter software has been designed to convert the ANFEAD finding aids and EAC-CPF r authority records into RDF datasets compliant with RiC-O. We were able to complete this project in 2019 with funding received from the Ministry of Culture, which supports initiatives of French cultural institutions wishing to explore semantic web possibilities. We worked in close association with Sparna and its founder, Thomas Francart, an expert in semantic technologies.

Using RiC-O Converter, we have now successfully converted all our ANF finding aides and authority records into RDF format datasets compliant with RiC-O. And, of course, we can carry out further conversions whenever necessary, for example when major updates occur in our set of metadata.

RiC-O Converter also brought to light a number of quality issues (lack of precision, bad use of the EAD format, etc.) in our metadata, of which we would probably have otherwise remained unaware. Work on the conversion project was therefore a contributory factor in drawing up an overall quality enhancement programme.

We have now also embarked on a process of fleshing out and standardising our authority records and reference data ( 3) as a means of pooling all the contextual descriptive data mentioned earlier.

Will everybody have access to the RiC-O Converter?

RiC-O Converter is very easy to use, even for the uninitiated.

We have developed this system for the ANF, while also bearing in mind its potential uses for all archival institutions or any other bodies with EAD format finding aids and/or EAC-CPF recordsl interested in using web data technologies, insofar as their metadata is compliant with the RiC standard. This is why we have opted for a free software. Anyone will therefore be able to use this software and to adapt its code for their own purposes.

What still has to be done before users will actually be affected by these changes?

RiC-O Converter is only one stage in a much bigger, more extensive process.

Schema Ric-0 EN

While it is good to have full sets of RDF archival metadata, it is still necessary to develop tools for performing queries and visualising this data. ANF does not yet have an interface of this type, even though we have started to consider our options to figure out the next stages in the process.

The ALEGORIA (http://alegoria.ign.fr/) research project, in which we are involved, develops an innovative search interface for aerial photo collections with RiC-O ontology as the main metadata reference model for this project. And we have also been able to draw inspiration from interfaces that are technically comparable, such as https://data.bnf.fr/, based on an ontology compliant with IFLA LRM.

We also need to consider how these semantic building blocks can be included in the existing ANF information system. One particular need is that of finding a means of updating RDF datasets when colleagues make changes to the original data in the systems they use in their work.

How can the international archiving community be of assistance?

First and foremost, by letting us know what it thinks of RiC!

In December 2019, ICA’s EGAD Expert Group published working versions of the conceptual model and ontology, inviting comments from the archiving community as a prelude to launching further projects.

The EGAD working party is planning to publish Version 1.0 of RiC-CM and RiC-O this summer, to give reference status to these comprehensive stable versions developed in conjunction with the archival community. It goes without saying that RiC will continue to evolve. We nevertheless hope that these standards will form the basis for major projects, such as portals, which will probably be the case with the international SNAC portal (Social Networks and Archival Context.)

Florence Clavaud, General curator, head of the Office of Authority records and Vocabularies at the French National Archives Executive member of ICA EGAD Lead of RiC-O development team

Note (1) for example, IFLA LRM, the conceptual bibliographical reference model published in 2017: see https://www.ifla.org/publications/node/11412 (return to text)

Note (2) EAD is currently in widespread use throughout the world, especially in French archives and as an exchange format for portals such as Archives Portal Europe. This DTD or schema defines the vocabulary and the rules to apply for producing digital archival finding aids according to ICA ISAD(G) and using XML (return to text)

Note (3) Controlled vocabularies, information about people and places, etc. (return to text)

RESOURCES AND INFORMATION

About RiC
Latest news about the development of Records in Contexts https://www.ica.org/en/news-on-records-in-contexts-standard

About RiC-O Converter