Community Call/05

From Wikibase Personal data
Jump to navigation Jump to search

Invitation

The invitation, including the Zoom link, is here.

Agenda

  • Goal today: best practices for working together, not actually solving the problems
  • Communication channels and rationale
  • Who reads from the wiki?
  • hence: need for an ontology
    • How do we structure Internet of Things (IoT) devices like Tesla Cars or Amazon Echo / Alexa?
    • Should we make on item by model? Group of model?
    • One item by services? But what is the services? Typically Amazon Echo devices are linked to Alexa services, but they may differ in data collection.
    • Can we do a type of data collection and inject it in all Tesla Cars Model?
    • How do we index data controllers? Etc?
    • Project:Ontology
  • Overview of the wiki base technology. [video timecode 39:41]
  • Help:Templates
    • Including importantly how to enable a new one
    • Template syntax (maybe that might need to be a separate call)
  • Help:SPARQL
    • Querying .. whatever the equivalent of SQL is
    • How to generate visual query results (nodes, tables etc)
  • Merging?
  • Update queries/batch operations Help:Quick Statements
  • Other cool/useful stuff you can do that I might not be aware of (edited).>
USEFUL TOOL: Descriptions: Show the description of items and properties when hovering them. https://www.wikidata.org/wiki/MediaWiki:Gadget-Descriptions.js

Final video

Key timepoints and notes

  • [15:00] Explanation of how the GDPR request generator template is loaded:
  • [25:25] Explanation of User's own common.js
  • [28:28] Explanation of Gadgets MediaWiki:Gadgets-definition
  • [39:41] Explanation of Wikibase Technology: SQL datastore of RDF triples. There is a CRUD API that interfaces with that. But it also uses a graph database to enable SPARQL queries. The SQL database gets fed into BlazeGraph using the 'updater' but this has a delay (a second to several hours). So query results may not be as up to date as what is in the database.
    • [42:05] how to access SPARQL query service
    • Help:Wikibase_technology
    • [43:00] API [[API help page]]
      • [43:12] API Sandbox
      • The moment you make an HTTP Post request via the API, that goes into the wiki's SQL database.
      • [43:40] Explanation of what a "claim" is. - it is effectively an RDF triple. e.g. Item - Property - Item or Item - Property - String/datatype
      • Wikibase has a separate layer, its interpretation of the underlying RDF triples.
  • [46:29] Namespaces
    • [48:55] Can you move pages between Namespaces? Yes. (show how)
    • [49:25] Do users have their own sandbox namespace? Yes everything under your username is a namespace e.g. you can just make User:Alexbfree/ExampleSandboxPage. (This could be useful for testing things privately before they are shared with other users)
    • [50:46] The Easy Query Gadget MediaWiki:Gadget-EasyQuery.js which shows all items that have this property. Activated by default for everyone. Just click the [...] next to a title.
      • [51:25] Using the Easy Query Gadget to see a quick node graph for an Item, and then access the SPARQL for that query.
    • [52:14] Show Derived Statements gadget MediaWiki:Gadget-relateditems.js - see which pages link to/use this one.
  • [53:39] SPARQL demonstrations
    • still need to add timecodes/notes here...