h1. OpenAtlas INDIGO Vocabs Meeting 2022-03-31 Updated information in the course of the meeting is %{color:green}in color%. Every participant is invited to add and adapt. h2. Participants * *OpenAtlas*: Alexander Watzinger, Bernhard Koschiček-Krombholz * *LBI ArchPro*: Geert Verhoeven, Jona Schlegel * *ARCHE/Vocabs*: Massimiliano Carloni This meeting is about importing INDIGO thesaurus from ACDH-CH "Vocabs":https://vocabs.acdh.oeaw.ac.at into OpenAtlas (#1663) h3. Import * OpenAtlas can link to external references (e.g. Wikidata) with SKOS match *exact* and *close* but not the used *broad*, *broader*, *related* and *narrower* matches. * There seems to be some issues with linking to vocabs (e.g. id format "0030") -> %{color:green}Comes from the CSV import. We will further discuss alternatives like using terms instead of numbers which will also make for more readable URLs% * Decide about multiple/single selection and what to link with. ** %{color:green}Graffito Creator (Single) -> Person and Group% ** %{color:green}Graffito Class (Single) -> Artifact% ** %{color:green}Graffito Component (Multiple) -> Artifact% ** %{color:green}Graffito Type (Multiple) -> Artifact% ** %{color:green}Community Term -> will not be imported to OpenAtlas% * Is the link to Vocabs enough or do we want to track related information in the description field for: ** Alternative names (e.g. translations) -> %{color:green}Link to Vocabs% ** Dublin Core information, e.g. for Background: http://purl.org/dc/terms/source: [{'@value': 'van Treeck 2001'}, {'@value': 'McDonald 2013'}] %{color:green}-> Link to Vocabs. Question: is it possible to link a reference to a type? Not currently but we can think about implementing but have to consult the CIDOC CRM model first. Please create a Redmine issue if this is needed.% ** Are there also descriptions in Vocabs? %{color:green}-> Descriptions are possible in Vocabs but weren't used yet% h3. Synchronization As expected automatic synchronization between OpenAtlas and Vocabs isn't possible for various reasons. Most obvious workflows would be: * Freeze Vocabs and import to OpenAtlas before work on data starts * Keep both systems in sync manually %{color:green}-> will be the most likely workflow. In summer 2022 will be a first version with a fixed structure and it is assumed that only a small amount of terms may be added later.%