User:John Cummings/hub

To do

• Consolodate sections
• Ask leads whats missing

Needs assessment

A partnership word cloud

Needs assessments research results

During the interviews we have identified five key themes in the needs identified by participants and the support they requested. The five themes that match needs and wanted support are around Funding and staff, Documentation, Management and communication, Support for collaboration and Technical skills and tools. This is the basis for future working groups and perhaps for some kind of start for a global partnership program. We would also like to take this opportunity to thank everyone who participated in the interviews and discussions! Thank you!

Results and summaries

A partnership word cloud

Results and summaries are reported continuously on the page for interviews. You can already see clear patterns and needs that we will continue to work with.

One such question is how to create a network that is broad enough and open to many issues, but still specialized to be able to solve more difficult, specific issues. It may be that you are looking for special knowledge that some people possess. How to create a network where the focus is on helping each other?

Read more at Needs assessment/interviews. Tips and suggestions are still highly welcome!

Contact tore.danielssonwikimedia.se and continue the discussion.

Needs assessment – video recorded interviews

Previous interviews have shown quite similar needs and where more resources are needed. We have now switched over and started recording interviews to be able to go back and follow longer reasoning on certain issues. Examples of questions we've discussed - Do we need more Wikimedia ambassadors to talk about the benefits of the platforms? How could one conduct courses and collaborations with a global perspective? What would a mentorship program for content partnership look like?

Do you have any more ideas? Feel free to get in contact with: tore.danielssonwikimedia.se

Needs assessment interviews

We have started to do interviews with Chapters and User groups to discuss needs and what possible support could look like in some of the work of the Content Partnerships Hub. It has been very rewarding and interesting conversations. We are grateful for the generous attitude we meet from colleagues around the world. There is a great interest in working more together internationally and helping each other where there are similar issues that need to be resolved. The discussions are gradually reported on Needs assessment interviews.

The discussions have been based on some questions around partnerships and we want as many as possible to give their views on how the area can be developed. If you have time for a conversation, feel free to send an email to tore.danielssonwikimedia.se

International partnerships

Working with partners to make content available

United Nations Environment Programme Food Waste Index Report 2021

One central part of the work with the Content Partnerships Hub is the collaboration with International Government Organizations (IGOs) and International Non-Governmental Organizations (INGOs) to make content available on the Wikimedia platforms. We are talking with several UN agencies, for example, around opening up their content to free licenses, and finding best ways for how to get them to the Wikimedia platforms. So far, the UN Environmental Programme has released their flagship publications under a Wikimedia Commons-compatible license, which means that these publications can be used to write or illustrate articles, for example.

IGO/INGO

United Nations Environment Programme have shared even more of their publications on Wikimedia Commons, please take a look and use these instructions to add text from them into Wikipedia. Let us know what you've done with the content!

United Nations Environment Programme Drowning in Plastics – Marine Litter and Plastic Waste Vital Graphics, one of the publications.

UN Environment Programme sharing their knowledge on Wikipedia

United Nations Environment Programme Drowning in Plastics – Marine Litter and Plastic Waste Vital Graphics, one of the publications.

We have been working with the UN Environement Programme to share their knowledge on Wikipedia, they have released their 9 flagship publications under an open license so we can use the text within Wikipedia articles. So far their text has been added to articles which are seen over 600,000 times per month including Plastic pollution, Microplastics, Marine plastic pollution, Soil, Agriculture, Fishing, Aquaculture and Natural disaster. Please add text from their publications on Commons into Wikipedia using these instructions.

Encyclopaedia of Life release their 2 million species descriptions under CC0

Wikimedia | © OpenStreetMap

Encyclopaedia of Life have released their 2 million species descriptions under CC0 and are working with us to import them into Wikipedia and Wikidata. We are building tools and prototypes to make use of this content, to date we have created:

• A working prototype of creating a complete article from data from Encyclopaedia of Life including infoboxes, and automatic image lookup from Wikidata.
• An article for every plant species in Malta, over 300 articles have been created using data and descriptions from Encyclopedia of Life including infoboxes with images.
• Working prototypes of an interactive map to visualise species distribution data from any source, currently taken from Plants of the World Online. Example 1, example 2

Software

Wrapping up some ISA-things

Only love for the ISA tool

A few things that we had been working on with ISA were finally deployed.

Translating the interface has been possible for a while now, using Translatewiki. However, these translations were not automatically deployed and therefore not presented to the user. Deploying these can be done manually by one of the maintainers, but this is something that can easily be forgotten or cause issues. Now there is a script that runs automatically once a day and updates the translation files.

Users are now referenced by id rather than username. This quite a large change in the backend even though it may not look any different to the user. There was at least one bug that was caused by this where you couldn't access a campaign that you had created if your username was too long. It also means that the database will take less space because every reference to a user is just a number rather than a string of characters.

You can now choose which language to use when searching for what an image depicts. Previously the language of the interface was used for this. While more languages have been added thorough translations there are still only a handful of them. The new setting allows you to either select a specific language or use the same one as the interface.

If you want to see a full log of what's been done, including links to Phabricator tasks with more technical information, head over to m:Content_Partnerships_Hub/Software/ISA_Tool.

Improving ISA

ISA participate page mock

ISA participate page mock

ISA is a tool that allows you to add structured data to images on Wikimedia Commons. It was developed in collaboration between Wiki In Africa, Histropedia and the Structured Data on Commons project. While it is still running and being used frequently, it has seen few updates since the initial development project ended.

We will be helping out with some development time to make sure that the tool will continue to be usable. So far we have solved a couple of security issues. We have now started working on some of the more noticeable issues, including slow loading of campaign pages and localisation, with more features planned to come.

Helping with getting Pattypan back on track

Pattypan

Pattypan is a tool for uploading large sets of images. It is used widely by GLAM institutions to contribute to Wikimedia Commons.

Last year Pattypan stopped working which caused a lot of problems and delays. The issue stemmed from a change in the MediaWiki API. This affected a software library that Pattypan, among other tools, is using. At the same time Pattypan only supported a version of Java that was not compatible with some dependencies. This meant that the process of running and testing new versions of Pattypan was difficult.

We decided to see if there was something that we could help with to get Pattypan up and running again. Since it was requested by Albin Larsson, the volunteer developer working on the issue, we worked on packaging the tool. This means that the end user doesn't need to download any extra things besides the tool itself.

After some work, testing and several experimental releases a new version of Pattypan was released February 7.

Strategic data uploads

Adding more SDC statements

Since September 2021, Wikimedia Sverige has been adding SDC statements to photos from Wiki Loves Monuments around the world. You can find a summary of the project and a list of countries we have worked on on this page.

The last photo set we worked with in December was India, thanks to the tip from Bodhisattwa. A country of prolific cultural heritage photographers and one of our largest photo sets to date.

This means that we ended the year with a significant milestone – over 500 000 processed WLM photos and over 1 million new SDC statements. We are very happy to have contributed to the work done by, among others, Multichill and his bot.

SDC for fun and profit: detecting bad coordinates

Structured Data on Commons can help you analyse information about media files and spot inconsistencies and errors. One example is geographical coordinates.

Many photographs on Commons contain the coordinates of where they were taken in their metadata, as saved by the device (mobile phone or camera). This is very useful information, but it's not always correct. If your camera saves erroneous coordinates with your photo, there's a risk you won't notice it when uploading the photo to Commons.

Then a bot comes and automatically copies the coordinates to SDC as coordinates of the point of view. Now the error is in two places!

But now that the coordinates are included in SDC, it also makes them easier to spot. We can use the Wikimedia Commons Query Service to plot photos on a map, and anything suspicious will stand out.

We discovered this when working with structured data in Wiki Loves Monuments photos. First, we add information about the competition to all photos in a country's category, for example Images from Wiki Loves Monuments 2021 in Iran.

Then we can write a query to put all photos participating in WLM Iran on a map.

We can easily see all the files whose coordinates are clearly not in Iran! We can now remove the coordinates from both the file page's wikitext and the SDC.

Note that the Wikimedia Commons Query Service is still in beta, and doesn't update automatically. The data is updated via weekly dumps. So you won't be able to see your edits reflected in the map until after a couple days.

You can follow our work with SDC and WLM photos here. There you can find a list of the countries we've worked on – some recent updates include Georgia and Venezuela – as well links to the SPARQL map queries.

How many bad coordinates can you spot in your country?

Cultural heritage on Wikidata – thousands of monuments in Norway

Some of the cultural heritage monuments in Norway – there are over 17,000 on Wikidata!

We created almost 14,000 items on Wikidata for cultural heritage monuments in Norway. The data for the items was taken from the Monuments database. These were originally from the Norwegian Riksantikvaren’s site] and include the ID in their database: Kulturminne ID (P758).

An attempt to match the items against already existing Wikidata items was made, to add the ID to them. It was however hard to get the correct items, as the articles for many of the monuments were about wider concepts. E.g. a piece of rock art linking to a village where it is located.

Items with the Kulturminne ID property, including the newly created ones, can be viewed in this query.

Structured Data uploads continue

We have set up a page in the hub portal with a summary of our data uploads, including the work with SDC and Wiki Loves Monuments. There you can find an overview of the countries we have worked with. We are now up to a dozen countries, with Norway and Tunisia as the most recent additions.

Tools and data – WMSE at the WikidataCon 2021

We had a great opportunity to talk about different aspects of the hub initiative at WikidataCon 2021.

Alicia Fagerving participated in the panel Please don’t get hit by a bus! Towards a resilient and sustainable Wikidata tool ecosystem, moderated by Sandra Fauconnier. Several developers exchanged their experiences of creating and maintaining Wikimedia tools, with all the joys and frustrations it involves. Tools, such as Pattypan and OpenRefine, are the key to working with big amounts with data and other resources, so it's definitely something we're keeping an eye on as an emerging Content Partnerships Hub.

Alicia also talked about our work with Structured Data on Commons and Wiki Loves Monuments photos.

Once the recordings of these talks are available, you will be able to find them here.

SDC for Wiki Loves Monuments

One of the areas the Hub team has been involved in are strategic uploads of data, including Structured Data on Commons.

Over September and early October, we worked with photos from the Wiki Loves Monuments competitions around the world. We added data on what the photos depict and which edition of the competition they participated in. You can read more about our workflow in Wikimedia Sverige's White Paper on SDC, which was published earlier this year after our pilot uploads.

This fall so far, we have updated over 188,000 photos with 498,000 new statements. See an example here.

The countries we've worked with include Australia, Bolivia, Chile, Colombia, Finland, Malta and Serbia. The workflow of adding the statements looks like this:

• 
  The contents of the Wiki Loves Monuments template are converted to participant in statements.
• 
  The contents of the Heritage site in Israel template are converted to depicts and location of creation statements after locating the monument on Wikidata.

Why are we doing this? Structured Data on Commons makes it possible to learn more about the photos uploaded by WLM participants. The data we've been adding is a complement to the data that have been added by other community members, such as Multichill, who has been converting the information about the authors of the photos and their geographical coordinates into structured data statements. With all these puzzle pieces together, you can use the Wikimedia Commons Query Service to find all the photos that participated in Wiki Loves Monuments in a certain country, and plot them on a map if they have coordinates.

Here's such a query for WLM in Malta, with the photos color-coded by the year of the competition.

Wiki Loves Monuments on Wikidata

In order to tag photos on Commons with the WLM edition they participated in, we first need to create Wikidata items for the editions!

This had not been done in a systematic way before, so it was something we had to include in our workflow. And here's the result – the Wikidata items of the WLM editions in several countries. The list will be updated automatically when new items are created.

Other

Creating a visual identity – compete with your logo designs

As the hub is continuing to grow and form, we are working on the graphic identity. An important part of this is the logo, and that's where you come in!

We are running a competition to find a new logo: the winner will pick their prize as an item of their choice from the Wikipedia Store.

Please share your ideas and designs here by April 30th.

Looking forward to seeing your designs!