Research Observatory: ISO 25964 - Interoperability in the Context of Vocabulary Mapping and Interchange

ISO 25964 Parts 2 – Interoperability in the Context of Vocabulary Mapping and Interchange

ISO 25964 Part 2 covers interoperability between different thesauri and with other types of structured vocabulary. It is essentially a primer on how to map concepts. This talk discusses interoperability in the context of vocabulary mapping and interchange. Vocabulary mapping has always been an opportunity to facilitate more comprehensive and exhaustive information retrieval before the Internet, for Web search, and most recently for large language model (LLM) AI. There are parallels between Web search adoption and the effective use of LLMs. The approaches to integrating knowledge organization systems (KOS) into LLM prompts are the same as for integrating KOS in information retrieval applications.

This talk lays out the history of the quest for an RDF interchange format for networked KOS and alignment of ISO 25964 with the W3C Simple Knowledge Organization System (SKOS) recommendations. The first Networked Knowledge Organization Systems (NKOS) workshop was organized as part of the second ACM Conference oi.e. n Digital Libraries held in Philadelphia, PA in 1997. SKOS was developed as an RDF thesaurus interchange format beginning in 1997. It was published as a W3C recommendation in 2009. These two initiatives were not just parallel, they were inter-related.

Today, obtaining a common vocabulary remains a primary challenge in building modern data and information management applications. LLMs do not make the vocabulary problem go away. The new AI is burdened with the weight of ambiguity reflected in the un-curated collection that includes everything and anything. The challenge is how language tools like taxonomies, thesauri, ontologies, and even simple lists of synonyms can be incorporated into the content processing workflows to support more effective AI prompting and machine interpretation, so they can be activated as language processing agents to clarify, refine, and define the meaning of content and user intent.