This week I presented recent work on Paradicms at the 2023 LD4 Conference on Linked Data.
End-to-end Linked Data with minimal computing
Systems that incorporate Linked Data often have many moving parts, such as graph databases, extract-transform-load pipelines, SPARQL endpoints, and web applications. Complexity makes the systems hard to sustain: they are brittle, expensive, and require frequent upkeep by people with advanced technical skills.
It doesn’t have to be that way. This presentation will show how the Paradicms framework (https://paradicms.github.io) leverages commodity tools (spreadsheets, Markdown), freemium services (Airtable, GitHub), Semantic Web technologies (JSON-LD, RDF/JS), existing knowledge bases (Wikidata), and static website generation, making it possible to build lightweight workflows and applications that show off the advantages of end-to-end Linked Data without most of the baggage.
This approach de-emphasizes the laborious aspects of data models, APIs, and serialization formats, on the assumption that few end users are as interested in the underlying technologies. Instead it seeks to answer the question, “What can those technologies do for the user?” and explores ways in which Linked Data can empower user interfaces, from faceted and fulltext search to digital exhibitions and beyond.
This presentation will provide a high-level overview of how the components of Paradicms work together, demonstrated through a series of web applications developed using the framework.