Technology

A gradually growing definition of the site's technological design challenges and decisions.

General

Wikidocumentaries is a project, which aggregates content from Wikimedia projects, openly licensed GLAM archives through queries and from the visitors. The project allows less notable topics, such as biographies or local history.

The first selection of topics will cover the geographical area of Finland, people related to the the country, events in Finland etc. The topics are retrieved from Wikidata and stored in Wikidocumentaries, which is also a Wikibase. The selection of topics may also be created organically on the go.

Content modules

Each content page consists of content modules that are added to the page depending on the page type, topic type and available content. The list presents ideas for the modules, not everything is created immediately. Each module is described in more detail on their own page.

Modules can have different formats, based on the same content. They can be either

Page modules, when they are part of a topic page or the landing page
Linkboxes, when they are added as related info or teasers. There may be need for different formats.
Module pages, when the module is expanded to use a full page

See full list of Content modules

Reading content through APIs

Content is primarily not stored in Wikidocumentaries. It is read from organisations and platforms that offer open and openly licensed content for use through APIs. The structure of the API calls will be stored in Wikidocumentaries and the information stored about the topics will be used to construct the queries.

See the demo and the code.

Tools

Each topic can be complemented with more info, the relating article can be made better, images can be tagged with names, keywords or location etc. The site will host tools for working with the materials displayed.

See the Tools page

Using Wikibase

The site runs on a Wikibase, which stores a subset of Wikidata items. Wikidocumentaries will read Wikidata for new and changed info, and Wikidata will import data from Wikidocumentaries when it meets certain criteria.

Other data to store in the Wikibase are definitions of APIs in various repositories where Wikidocumentaries expects to find content. The technology of Wikidocumentaries is based on an experimental setup of using Wikidata data to create the backbone of the site information.

Data modelling

Items and properties
Topics, APIs

Questions to solve about data modelling for Wikidocumentaries

What is created locally for a topic that has a corresponding item in Wikidata?
Is there a way we can use Wikidata properties for items live, without needing to define them locally?

Creating topics in Wikibase

Page = item in Wikidata and/or Wikidocumentaries. Each topic in Wikidocumentaries has a page.

The initial set of topics in Wikidocumentaries are created by selecting a set of Wikidata items and creating them in Wikidocumentaries.
When the user navigates to a page that does not exist in Wikidocumentaries, but for which there is an item in Wikidata (through a link in Wikipedia text for example), the page is automatically created.
The user may create additional topics in Wikidocumentaries with a set of minimum required properties.

See more about Creating topics

Using ShapeExpressions

The scope of automatically created item may at some point be defined as a Shape Expression, which will probably be stored in the Federated Wikibases registry. This definition can help add whole thematic areas at once in a coordinated manner.
The newly created items in Wikidocumentaries can be automatically imported to Wikidata, when they meet a certain criteria. This can also be expressed as a ShEx.

Questions

Can Wikidata properties be used LIVE in Wikidocumentaries?
What can be a realistic schedule for deploying Shape Expressions?

User interface

Wikidocumentaries starts by creating the necessary functionalities in the MediaWiki interface. The goal is, however, to create an independent interface that reads the underlying Wikibase and MediaWiki information through an API.