People who are looking for information don’t think that much about how that information is stored. They just want to find the information quickly and efficiently, and to assess its quality and authenticity. With the World Wide Web, and Linked Data in particular, archives have the opportunity to move from siloed architectures towards data driven structures. For the archival field, Records in Contexts is the means to achieve that. This part of the Application Guidelines looks at the benefits of RiC for end users, archival practitioners and managers in archival institutions respectively. After reading this part you should know why it is worthwhile to invest in understanding and using RiC.
The benefit for end users comes mostly from the network structure that is one of the foundations of RiC. After all, records may not only be related through a common provenance, but may also be related to common people and places. This makes it easier for end users to answer questions like: ‘what archival material is available about my home?’ or ‘what archival material relates to my grandmother?’ or ‘what archival material relates to healthcare in the 1950s?’.
RiC ensures that multiple routes to information are created. Currently you need to know exactly where archived information is in order to find it, or you need to know who created the records – something that is not usually intuitive for a member of the public. Thanks to RiC’s network structure it is possible to use relationships to get to certain information independent of where it is held. With RiC based metadata, any RiC entity (such as a Person or a Place) can become an access point alongside the hierarchical structure of an original finding aid. So, in addition to drilling down from the fonds to the record, browsing via persons and places becomes much more effective as they are part of a network of data and documented relationships. That network could be between archival resources within one repository, but also linked to data somewhere else, outside of your institution. Related content may include, for example, collections by the same creator held at different institutions, collections by members of the same family held at the same or different institutions, or collections relating to the same event held at the same or different institutions.
Because multiple avenues of access are offered, more types of users can reach archival material and, depending what the data is linked to, even users who are not specifically looking for archival material can find it too.
Using Records in Contexts creates a network of data, which makes data better contextualized. The data is no longer presented in isolation (like the old silos of finding aids, indexes and image databases), but always related to other sources, people or places, which leads to a better understanding of the data. As a result, the data becomes more meaningful.
The network structure of RiC also enables new visualizations of search results. We’re already familiar with the table (when the metadata is most important), the timeline (when chronology is most important) and the map (when geography is most important). For the user who is interested in the relationships between the people, places, events and records that form the network, the representation (or visualization) of this network of related entities as a graph can be very helpful. It enables you to browse or explore such a network by following a trail of nodes that are of interest.
When you combine RiC with Linked Data techniques you can infer information that your local system doesn’t know. Linked Data is a way of expressing information in statements following the pattern ‘subject/predicate/object’, as in ‘William Shakespeare (subject) is the author of (predicate) Macbeth (object).’ RiC implemented as Linked Data adds benefits to resource discovery, in particular the ability to include data that is not stored locally but is available in a related record elsewhere. For example, another system might store the statement ‘Macbeth was published in 1623’. When linked to the entity ‘Macbeth’, both pieces of information–authorship and publication date of Macbeth–can become available to the same system despite originating in separate systems.
When RiC is combined with Linked Data techniques it is ideally suited for:
selecting data: doing research with your own specific, dedicated subset
combining data: relating the research data with data available in other sources
inferring data: using the available, standardized rules in RiC-O or creating your own rules to add data that can be deduced from the existing metadata
keeping track of references: storing the origin of the data for future researchers to study
Although these activities can be carried out using existing techniques, Linked Data makes it possible to perform them more precisely and will provide more data for the same or less effort. Increasingly, researchers are looking for technical ways to use archival sources, and RiC enables these approaches to be used alongside the traditional methods. As a result, more people will be able to find and use your archival material.
Archival practitioners are passionate about making sure that the quality of their metadata is high. However, practitioners often encounter the limits of the existing data model. For example, within a traditional hierarchical system, a record can be described at only one position in the hierarchy. Most of the time this position is related to the arrangement and the provenance of the record (e.g., financial records are part of the series of records created by the financial department), and not accessible through the subject (e.g. ‘travel expenses’ or ‘personnel’) that an end user might want to study. That means an end user must know how to drill down the hierarchy to discover the right archival source. With Records in Contexts (note that ‘Contexts’ is plural!) subjects can be added at the same time as the provenance and the arrangement are added. RiC does not consider these two elements of context as the only two, as is the case in a hierarchy, and allows professionals to move from a mono-hierarchical perspective to a multidimensional description. They no longer have to choose a single location in a hierarchy for the records they describe, but rather are invited to represent the network of entities in which these records are placed.
With born-digital records this problem becomes even more complex: in a document management system there are all kinds of relations between records and computer files. Records in Contexts defines clearcut entities, and relations to connect those entities to each other. These relations give several points of entry to a record, and also essential contexts. In a hierarchical system much of this information would get lost. At best, it is stored in the description, which makes it messy and difficult to analyse.
Another important benefit is that Records in Contexts introduces new concepts to model digital and digitised sources. An archival record can be instantiated more than once, besides its original paper form or digitally appraised computer file. It makes the conceptual model robust for the future, while enabling, for instance, all kinds of AI technology to be used as well. Handwritten Text Recognition (HTR) and Optical Character Recognition are tremendously useful to all kinds of users: you can search within the data of the record itself. Because in Records in Contexts the transcript is a separate entity, you can store metadata about it. When was the transcript made, from which scan, and with which model? When you have a better HTR model, you can easily select old versions of the texts to be replaced by new ones. See §6a.3.4 for more on this.
The introduction of additional entities like Activity and Rule (see §6a.3.2) provides for nuanced expressions of archival practice, helping the archivist to model clearly the provenance and the history of archival material. Such approaches enable a more unified approach to describing records before and after their transfer to an archival institution. Metadata captured by records creators can be reused in archival description systems. Simply put: Records in Contexts makes sure that metadata is captured where it should be. With more entities, and properties for those entities, and relations with a specific name, the practitioner has many more options to describe records in their contexts.
When you combine RiC with Linked Data you have the opportunity to integrate all kinds of controlled vocabularies, authority records and value lists into your system, and to use them to standardize metadata. Controlled value lists are already in use in the archival field to varying degrees, but the benefit of the Linked Data technique is that it is so much more interoperable with other sources. As an archivist you can use this to benefit from other people’s work. For instance, if a certain agent is important to your collection, you probably want to include the name of this person in one of your own lists. You can link this list to external sources, such as Wikidata. When someone else has found out the date of birth and death, you can simply import this information. It saves you the time of researching those dates yourself. And it prevents redundancy of metadata. The more you standardize, the more time it saves you. This leads to improved work processes.
Introducing RiC as Linked Data can be realized by using the standardized RiC Ontology (RiC-O). RiC-O contains all the RiC entities and its predefined types of properties, called ‘relations’, complemented by a lot of other entities and properties. Some of these enable you to provide more accurate descriptions, others enable backwards compatibility. Using RiC-O to publish your archival metadata helps to link it with archival description in other heritage institutions and exchange metadata if necessary, widening the network of data available to the researcher.
With Records in Contexts we have one global and consistent reference model, maintained by and for the professional community which uses it. It can be used as a way to break down internal silos and design a data-oriented enterprise architecture. In this architecture, processes can be organized more efficiently using software components that are interoperable and easily interchangeable - if the components comply with the standard. The resulting architecture is better suited to archival processes and will meet the needs of existing and future end users.
It is important to note that the standard can be introduced step-by-step following an iterative approach. The implementation can be organized as a series of modular projects carried out over time. During these projects the reference model can be extended if needed. For more on this see §3.
Introducing Linked Data and the Records in Contexts Ontology helps transform the metadata you hold and publish it in a variety of forms suitable for various needs. It also ensures your organization complies with international standards for reusable data sets in scientific research, like FAIR.
The introduction of Records in Contexts in your organization is worth the extra effort and resources as it results in greater control of the ever-increasing amount of data. Through a better understanding of the components of the data and the software, the quality of the metadata in your collection improves, as well as the efficiency of the processes handling it.
Records in Contexts makes your organization ready for the future. New technologies, like AI, can be hooked into the new infrastructure. For the first time it provides a framework for describing both born-digital material and analogue records, and as a result will help with the processing of both paper and digital records, as well as hybrid collections, which have grown in recent years. Thirdly, as a common reference model, Records in Contexts enables connection of your archival material and data to other organizations, especially if you implement it in Linked Data, and makes cooperation with these organizations easier. As RiC is adopted by more organizations, it is likely that funding partners will require you to introduce RiC into your institution and project, as a condition for funding.
The transformation of your finding aids from the old standards to Records in Contexts is a considerable undertaking, and will mean changes to how you work, but it will improve processes and enable researchers to use your data in ways that are not possible now. The change will not happen overnight. Implementing RiC is an iterative process and approaches to this, along with how to get started, will be described in the following parts.
