| No
|
Name
|
Task
|
Brief description
|
URL
|
Showcase presentation notes
|
Team
|
Development status
|
Audience
|
Slides
|
| 1
|
PagePile Visual Filter
|
phab:T230598
|
Auxiliary tool for PagePile to visually select a subset of a pile, in a manner similar to the Cat-a-lot gadget, and save it as a new pile.
|
https://tools.wmflabs.org/pagepile-visual-filter
|
Structured Data on Commons lets you declare which figure skater is shown in a file, and in which pose. But selecting all the images in a certain pose is hard if the figure skater’s category is split, e. g. by year. PetScan can collect all the files from the subcategories, but not let you select the right ones. You can export the result as a PagePile and feed that into the new PagePile Visual Filter tool, where you click on the images which show that pose, then export the result into a new PagePile. This one can then be fed into other tools like AC/DC (add structured data statements) or QuickCategories (add categories).
|
@LucasWerkmeister, @Harmonia_Amanda
|
development
|
Editors
|
File:Wikimania_Hackathon_2019_showcase_–_PagePile_Visual_Filter.pdf
|
| 2
|
Item creation helper for Wikidata
|
phab:T230599
|
Some concepts need several items on Wikidata to describe them, each time created the same way. For example, for an annual iteration of an figure skating competition, we need a general item for the competition, one for each of the ladies/men/couple/dance part, each of them further having two further items for the technical and free programs. This tool help create all these items, and link them together, with a simple form. Furthermore, it is a proof of concept for similar forms for other similar needs.
|
https://www.wikidata.org/wiki/User:Ash_Crow/creation-helpers
|
Gadget for Wikidata that allows you to create a group of statements (includes references), and then creates a batch of new items related to each other.
|
@Ash_Crow, @Harmonia_Amanda
|
implementation
|
Editors, Developers
|
File:Wikimania_2019_creation_helper.pdf
|
| 3
|
Knowledge crystals / Wikibase for structured discussions
|
phab:T230558
|
We want to structure discussions and debates so that they can crystallise knowledge. Each comment has its own wikibase item and connects to other items to form a network that represents the discussion. We set up the wikibase and created visualisations and a visual editor. Next, we want to find good use cases such as factual debate on Wikipedia's Talk pages.
|
http://dev.tietokide.fi/ and URL2: http://dev.tietokide.fi/?Q10
|
Organizing discussions by linking posts as attacking/agreeing with other arguments. Structure is based on argumentation theory, now implemented on Wikibase. Can hide irrelevant or less important details, to highligh most important arguments so that the participants can focus on those. The demonstration tool works, looking for collaborators who might use it.
|
@Jtuom @TuukkaH
|
development
|
Readers, Editors
|
|
| 4
|
Mortar: Interactive documentaion for GLAM Uploads
|
phab:T223608
|
Create and document a unified, user-friendly workflow for GLAM uploads on both Wikidata and Commons. We created a workflow using Pattypan and OpenRefine, defined the missing steps needed, and started an easily translatable and updatable documentation.
|
Tool link: https://tools.wmflabs.org/mortar/?lang=en
|
The general idea is that we have many tools to do some parts of importing items collections and photographs of these items on Wikidata and Commons. Wanted to see how we can add some mortar to ease the process and link the different tools toghether. Two basic use cases: Institution want to upload their items, or volunteer who takes own photos and wants to upload them. Implemented in node.js on toolforge. Can select the kind of item, e.g. picture; then upload. Second part is [?]. This is Mh-3110's first tool. (!!)
|
@Lena, @Anne-LaureM, @Ecritures, @Ash_Crow, @Mh-3110, @PierreSelim
|
Development
|
Readers, Editors
|
Slides: https://docs.google.com/presentation/d/1ANtFM7wGGCelSr7tUYv4bp_pyG6SwCHi1BDqNFtbEj0/edit?usp=sharing
|
| 5
|
iNaturalist Import Tool for Commons
|
phab:T222648
|
A UserScript that allows you to browse and import free-license images of a plant or animal from iNaturalist from within the corresponding Commons category or gallery page
|
Load a page like https://commons.wikimedia.org/wiki/Category:Tauraco_corythaix with this User Script installed: https://commons.wikimedia.org/wiki/User:Kaldari/inat2commons.js http://inaturalist.org
|
iNaturalist ( https://www.inaturalist.org/ ) explanation: project that is similar to WM projects, in that it's open-source crowdbuilt database of plants and animals; take pictures in your area, upload them to central DB, people can ID them and comment. Has a nice interface. They encourage the use of free licenses, and have 25 million of photos of plants and animals. Wrote user script to allow user to upload inaturalist photos to commons. [kaldari can't remember his password :-D] [bell rings] Once installed, you can go to any category or gallery page, and you will have an upload link that will allow you to select inaturalist photos that match. It is then uploaded and yay!
|
@kaldari
|
testing
|
Commons Editors
|
|
| 6
|
Scribe
|
phab:T230613
|
Supporting editors of low resource languages to create new articles on their Wikipedia, focusing on mobile devices
|
https://www.youtube.com/watch?v=-XZj0h5-hV0
|
Scribe on mobile is a tool that supports new editors on unresourced Wikipedias to write new articles. On mobile, you select a structure for an article. you select some references [lags] it suggests then building blocks to create an article.
|
Lucie-Aimée Kaffee, Hady Elsahar, Joe Reeve
|
development
|
(new) Editors
|
|
| 7
|
Parliament diagram tool
|
|
Allowing users to re-use previous party lists for new diagrams
|
https://tools.wmflabs.org/parliamentdiagram/parlitest.php
|
Based on experience volunteering on the Graphic lab. A previous tool was used to create PNG diagrams which were frequently requested to convert to SVG, and the tool that was developed a few years ago does this. A frequent request was to simplify the creation of diagrams. In this Hackathon, we allowed two new options: 1) Use Wikidata through this tool: select a party or multiple parties, and generate a list of elected in an assembly-shaped diagram in SVG. 2) Copy list of parties from a previous diagram. Now you can generale 20 diagrams very easily to cover the evolution of the parliament.
|
@slashme ( David Richfield) @ranjithsiji (Ranjith Siji) @Ambadyanands (Ambady Anand S)
|
maintenance
|
Readers, Editors
|
|
| 8
|
pywikisource
|
phab:T230529
|
Allowed Developers to get Wikisource related data like number of book pages, Current page quality status, proofreader, and validator in python.
|
https://github.com/indictechcom/pywikisource#Demo
|
There is a lot of activity on Wikisource in India, and they have a lot of requests to create tools for Wikisource related. Quality, proofreading, and validation data can be got by this python library.
|
@Jayprakash12345 @Suyash.dwivedi
|
implementation
|
Developers
|
|
| 9
|
Maps, maps, maps!
|
phab:T227036
|
We're making it easier to discover maps on Wikimedia Commons, and to better store their geospatial metadata!
|
https://github.com/bertspaan/wikimania-hackathon-2019
|
Finding maps on Commons is hard. You can use category:maps or rectified maps via Wikimaps Warper ( https://warper.wmflabs.org/ ). Georectifying uses fitted points on a map to show how the map fits over new maps. This can be used to drive much nicer discovery. This kind of tool should be on Commons too;and the metadata it uses should be available on WD/Commons. Maps are georectified, and the tool indexes the maps based on where they are in the world, and presents their outlines. 5 master repositories of maps have been used. The project created data specs, prototyped a service. Next is moving it to WMF labs, and to get the WD properties approved. Follow us on Github.
|
@bert, @JHeald, @thisismattmiller, @Abbe98
|
development
|
Editors, Developers
|
Slides: https://bertspaan.nl/wikimania-hackathon-2019/presentation
|
| 10
|
InternetArchiveBot
|
m:InternetArchiveBot
|
Better less demanding OAuth for privacy focused users, and UI enhancements.
|
https://tools.wmflabs.org/iabot
|
Internet Archive archives Webpages. It is really important to keep track of references on Wikipedia, and we use Internet Archive to do so. Making updates to admin interface. IABot requires lots of permissions to allow you to fix deadlinks; normally not needed. If you don't need to run the bot on your own, you can still queue up the bot, it will attribute the edits to your name. Changing Domain Data is time consuming and it can timeout on your browser. The tool allows you to make the queries on a separated server. If you want to actually run the tool, however, you can upgrade your permissions. UI has also been improved.
|
@cyberpower678, @skalman, @thedj
|
testing and enhancement
|
Editors and readers
|
|
| 11
|
Hostbot-AI Experiment Analysis
|
phab:T211160
|
Did an AI-powered version of enwiki's hostbot retain more newcomers than before.
|
http://isalix.hestia.feralhosting.com/hostbot-analysis.png
|
Teahouse is a place on English Wikipedia ( https://en.wikipedia.org/wiki/Wikipedia:Teahouse ) where newbies can get mentoring. The main way that people come there is that HostBot invites them. There are limited hosts on the Teahouse, so they only want 300 newcomers per day. The problem was solved by using heuristics like edit count to find suitable newbies to invite. Now we have ORES to determine the quality of people's edits ( https://www.mediawiki.org/wiki/ORES#Edit_quality ). We ran an A/B test to see the difference in how long people stay on WP depending on whether they're invited by heuristic-based or ORES-based inviter. AI bot was consistently better than heuristic-based.
|
@notconfusing
|
analysis
|
AI-developers / researchers/ mentors
|
|
| 12
|
Multilingual JavaScript
|
phab:T230665
|
A global project deserves a multilingual scripting/templating language.
|
https://docs.google.com/presentation/d/1DhFIYK6MmKwp0rLnSZ9M2uufzvx442BOa82dFvcn6yE/edit
|
Works for WMF, but this is a personal project. Although it’s likely that not everything will always be translated, knowledge of English should not be a prerequisite for participation in our projects. Right now English is used for writing Template code and Scribunto modules. But we could truely localize JavaScript and then use it to write templates. Rewrite all the strings and variable names and comments and method calls to translatethe code. Gives an option to have template code translated in various languages.
|
@cscott
|
design
|
Developers
|
|
| 13
|
Copyvio bot for Commons
|
phab:T230561 (see also phab:T31793)
|
A bot to help commons patrollers to find files require additional attention for copyright issues
|
https://commons.wikimedia.org/wiki/User:%D7%A2%D7%A8%D7%9F/Automated_copyright_check
|
To help editors and patrollers on Commons to find copyright problems. The bot goes over new files by new editors, without exif data, and creates a record. The record has a score based on a [] model that looks at the metadata of the file, and decides which need more attention. For example this new image, which is already flagged for speedy deletion. The metadata contains a string that indicates that it's from Facebook. This wasn't hard-coded. Google doesn't always scan FB, so a google search won't always find the duplicates. Hopefully this will help the commons community to find such copyvios.
|
@eranroz
|
development (POC)
|
Editors, Administrators
|
|
| 14
|
Various i18n tasks
|
phab:T229922
|
Fix a "lego" message to enable gramatically correct translations in Advanced Search. Took wikidata's FAQ page from 11% to 100% in Bulgarian
|
https://commons.wikimedia.org/wiki/File:I18n_lego_string.png
|
Worked on various small fixes. One example: Bug in Advanced Search Interface. Tag in search that says "sort by relevance"; it was two messages that were concatenated during runtime, and in some languages this isn't grammatically correct. This is now fixed. Call to audience: get into the Wikidata Help pages and get translating!
|
@Tonina_Zhelyazkova_WMDE
|
implementation
|
Editors, Developers
|
|
| 15
|
Mentorship tools
|
phab:T228210
|
Newcomer homepage <https://www.mediawiki.org/wiki/Special:Mylanguage/Growth/Personalized_first_day/Newcomer_homepage> automatically assign a mentor to all newcomers, but it doesn't allow to change for eg. wikicourses or similar. This feature allows users to change the mentor.
|
https://docs.google.com/presentation/d/1yh3rmpCJlrHR94ycOrBaDxUrO6thV6LyXk1mrTpAJvk/edit
|
The mentorship system created by the Growth team programmatically assigns mentors to all newbies. Problem is that the mentor is assigned automatically, but if you're on a course, you want your instructor. This tool allows the mentor to take a mentee specifically.
|
Team:@Urbanecm
|
testing
|
Editors
|
|
| 16
|
Lightweight tool to visualise Wikipedia article contest flow & results
|
phab:T230464
|
Article contests and edit-a-thons might be fun for participants, but for rest of us it's hard to get insight into what exactly happened during these events. For that we visualise the flow and results of these events in data story style to be presentable for general public as well as media in after event press releses etc. During the hackathon we created basic structure to gather and store the data need and made an example visualisation to be easily expanded for other visualisations and additional data.
|
https://infoaed.ee/vagavesi/
|
Article contest: if you look through the articles written for the contest you don't see it easily, need to go through all the articles. [?] Article contest with users and articles [?] Visualisation goes through all the users and articles and articles written during the time of the contest. Example visualisation: articles by length shown as bubble diagram, so don't have to go through pages and text to figure it out. Second visualisation is a bit more tricky: users and number of articles written by them. Some registered participants and other users affecting article, so visualisation to split these groups. Can be useful when setting up press releases.
|
@tramm
|
development
|
Readers, Developers
|
|
| 17
|
Category overview
|
phab:T230678
|
There are often lots of articles in a category but no easy way to get an overview of the articles at a glance. This tool is for having a quick overview of a category on the wiki. Right now it only works on enwiki but it's easy to make it work on other projects too.
|
https://niharika29.github.io/wikitopics/
|
Work as product manager on tools team but this isn't relevant to this. Quick overview of topic on WP? Are category pages useful? Lots of info there, but hard to get overview from it. Javascript tool to get info about categories and present it in a usable way. [Demo: "fire"] Shows article and page extract with pics. [Demo "goats"] same: now we know that there's a rent-a-goat company ( https://en.wikipedia.org/wiki/Rent_A_Goat ). Can make category pages more useful. Might at some point be converted into a user-script
|
@Niharika29
|
analysis / design / development
|
Readers
|
|
| 18
|
Bookmarklet for URL Shortener
|
phab:T222360#5420068
|
The URL shortener is a great tool for generating short links to wiki pages, but you have to navigate to a specific page, then copy and paste your URL in to use this tool, which is time consuming. This bookmarklet turns that into a one-click process, giving you your short URL instantly.
|
https://phabricator.wikimedia.org/T222360#5420068
|
Usually works on editing team on VisualEditor, but now something different: URL Shortener (example: https://w.wiki/7Ed ) takes any URL from a WM site, even etherpad; go to w.wiki, click "shorten" and get a short URL. Could be easier, though: make bookmarklet that makes an API request and presents it for you. If you're on etherpad or phab, you're redirected because they don't have access to your credentials.
|
@ESanders
|
implementation
|
Readers, Developers
|
|
| 19
|
Offline voice based Wikipedia
|
|
If you don't have internet and maybe struggle with text it might be nice to hear Wikipedia content anyway
|
|
Cool if you could have WP offline with just your voice. Built on Mozilla DeepSpeech ( https://github.com/mozilla/DeepSpeech ) and Common Voice ( https://voice.mozilla.org/en ) projects, powered by TensorFlow. [Demo] Hello Sweden, will you please go offline. Waitu waitu waitu waitu waitu OK 1231 Cornflakes.Here is what I heard you say: Cornflakes. Let me ask you [?] {And not much happens due to the curse of live demos.} Here is what I heard you say: Solar system Let me ask Wikipedia. Hang on. Let me see if Wikipedia can help you hang on. I do have one file. It is too small. Perhaps we should wait a few weeks and try later. [Demo fail, author is happy to show later.]
|
@dr0ptp4kt
|
analysis / design / development / testing
|
listeners / voiced (for now)
|
|
