Knowledge as a Service
Knowledge as a service (KaaS) is a term in the Wikimedia community first formally described in the 2017 Strategic Direction document, and is defined as:
Knowledge as a service: To serve our users, we will become a platform that serves open knowledge to the world across interfaces and communities. We will build tools for allies and partners to organize and exchange free knowledge beyond Wikimedia. Our infrastructure will enable us and others to collect and use different forms of free, trusted knowledge.[1]
Movement strategy
[edit]In the final set of Movement Strategy recommendations for 2018-2020, it was specifically mentioned in recommendation 9 to "Innovate in Free Knowledge":
We will innovate in different content formats, develop new software functionalities for Wikimedia projects, better integrate various tools in the editing experience, establish partnerships with other free knowledge projects, and improve our policies to include more diverse domains of knowledge and to deliver knowledge as a service for all.[2]
Concerns
[edit]- Do we have a common working definition of what KaaS means?
- What is the commitment of resources to such a goal?
- What is the process for deciding how we support (or remove support for) actual KaaS services?
Project-specific questions
[edit]- In what capacity should Commons serve as an image or video hosting service?
Implementations
[edit]Examples
[edit]- Wikimedia API - application programming interfaces
- Wikidata Query Service (WDQS) - a powerful way to query Wikidata using SPARQL
- Wikimedia Commons
- image hotlinking - we permit and promote the use of images directly from Wikimedia servers ("hotlinking"). When you view an image and click "Use this file on the web", we present the URL to use as well as some example HTML code for embedding the image on any web page.
- InstantCommons – allow the usage of any uploaded media file from Wikimedia Commons in any MediaWiki installation world-wide.
- Scholia – service that creates visual scholarly profiles for topics, people, organizations, species, chemicals, etc using bibliographic and other information in Wikidata.
- Wikibase Cloud – free hosting service by Wikimedia Deutschland
- Technical resources
- PAWS – cloud-based Jupyter Lab and Notebook environment for Python/Julia coding bots and scripts
- Insert more examples of KaaS
Caveats
[edit]There have been situations where we have put a limit on public-facing Wikimedia services in the KaaS domain.
- Maps Service API. In 2020, the public maps API (for embedding maps) was "discontinued" and restricted to use within whitelisted Wikimedia sites due to concerns about load.
- WikiCommons Query Service – WCQS was launched requiring authentication (and still does to this today).
- This was not the expected launch state for many folks expecting a service like Wikidata Query Service. Requiring login to use the SPARQL-based service was a call made by the (then) Search team over concerns about performance, which is understandable even if it is not popular.
- However, the open question is: what is the ongoing process to move it out of this authenticated, low usage, state and to a more generally available state? How might other solutions, such as providing an authentication method via API, help?
- Commons:SPARQL_query_service/API_endpoint
- Workaround solution at StackExchange – manually login via a web browser, extract the provided session token, and use it in subsequent API calls with the header parameter "wcqsSession=<token>"