Reuniting Without Erasing: Preserving Cultural Specificities in the Age of Transversal Research
The value of information is dependent on its usability, but also on contextualisation and correct interpretation of data, which are deeply rooted in language. The Canadian Heritage Information Network (CHIN) is responsible for aggregating unilingual heritage data—most often in English or French, but also in other dialects or languages. In order to do so, CHIN uses Linked Open Data which typically link provided values with thesauri elements based on term equivalence and context rather than on definition alignment.
As terms and concepts they represent are collapsed through aggregation, the distinction between how the provider uses the term and how the thesaurus defines it is obfuscated. As a result, valuable data might be misinterpreted when federated. This loss in granularity hinders the adequate contextualisation, representation, and understanding of data necessary for research, especially that on non-dominant language concepts and cultures.
This presentation will explore CHIN’s approach to document linguistic and cultural specificity through three attributes appended to vocabulary data in addition to language tags: a reference culture, a conceptual framework, and an example. It will focus on how each of these can either be documented by a provider or derived by the aggregator based on the data structure from which the value is extracted (language; provider domain of expertise and database field; record).
This record of users' frameworks allows access to that of others and relies on a co-documentation of concepts that offers a more granular information and research landscape as well as a better understanding of concepts by practitioners.