�Structured Data on Commons��User Stories/Epics – v.0.9
Jan. 2, 2018
Commons User Types
Viewers: mainly visit Commons to find those sweet free files. They most likely never edit. Within this type are Remixers (people who will download, edit, and remix the images in their own new art) and Embedders ((bloggers, reporters, etc. who embed unaltered work on other pages)
Casual Uploader: actively (on average at least once every month) upload one image at a time (which may or may not be their own). These are amateurs who probably take most photos with their phones.
Batch Uploader: uploads 20 or more images at a time using a batch upload tool. They are typically someone associated with a GLAM project, but not always.
Wikimedia enhancers: Users on various Wiki projects who search for images to use on those projects (Wikipedia, WikiVoyage, etc).
Photographers: pros or semi-pros who actively upload their own images to Commons. They are generally using DSLR/mirrorless cameras but may use phones in a pinch.
Editors: actively edit media data on Commons for the sake of accuracy, completeness, or maintaining site quality.
Curators: actively categorize, group, and label images to make things organized and easy to find. They may also be involved in picking featured, quality, and valued images.
Tool Builders: volunteer developers who write and release software to supplement Commons functionality or fill in functional gaps.
Admins: users who have special abilities to enforce the rules and primarily act as site police. They may or may not actively donate media.
User type overlap
Viewer /
Consumer
Casual
Uploader
Batch
Uploader
(inc. GLAMs)
Photographers
Editor
Curators
Admins
Tool
Builders
Remixers
Embedders
Wiki Enhancer
Wants vs. needs
For each user story, there’s a very deliberate word choice when it comes to something the user may want (because it’s cool and/or makes life easier), vs something they need in order to accomplish a fundamental task that is part of their workflow that will change once SD is implemented.
Hopefully, the distinction between the two types of desires helps us to start thinking about prioritization, which will be key when it’s time to arrange items in the roadmap.
Generic Stories�(apply to everyone)
[Saga] All File Info in One Place
As a user, I want to be able to view all the metadata associated with a file on one page.
Metadata for a media file may be stored in 3 different places:��a.) on the Commons instance of MediaWiki�b.) on the Commons instance of Wikibase�c.) WikiData��But all this information should be available on one web page
Users can see all directly relevant metadata for the file on a single File page in Commons.
Value / Rationale:
Condition of Satisfaction:
[Epic] MediaInfo data available on File page
[Parent Saga] All File Info in One Place
As a user, I want to be able to view all the MediaInfo metadata associated with a file on its File page.
Metadata for a media file may be stored in 3 different places:��a.) on the Commons instance of MediaWiki�b.) on the Commons instance of Wikibase�c.) WikiData��But all this information should be available on one web page
Users can see all directly relevant MediaInfo metadata for the file on a single File page in Commons.
Value / Rationale:
Condition of Satisfaction:
[Story] Metadata can be viewed in all available languages on the File page
[Parent Epic] All File Info in One Place
As a user, when I visit a File page for a media item, metadata should be available for viewing in all entered languages.
Curators, editors, and even some viewers want to see metadata available in multiple languages:��- to know how much is available for this file�- to know if a file should be edited more/is incomplete�- etc
The UI for each structured data item provides functionality to see the metadata in more than language.
Value / Rationale:
Condition of Satisfaction:
[Story] Wikidata links within MediaInfo data available on File page
[Parent Epic] All File Info in One Place
As a user, when I click on a MediaInfo entity that links to a Wikidata Item, I want to see at least a preview of that Wikidata info on the page I’m already on.
With Wikibase Federation, two instances of Wikibase can link to each others entities. This means that the new Wikibase@Commons can link to Wikidata items.
However, it’s a sub-optimal user experience if the user is taken to a brand new page when they click on the Wikidata link.
Users can click on a Wikidata link on the File page and open up an overlay/popup that displays basic data about that Wikidata entity.��There should be a link/option (perhaps within the popup) for the user to go to the Wikidata site to see the full entity info.
Value / Rationale:
Condition of Satisfaction:
[Saga] “Effective” search
As a user, I want to be able to find media that I’m looking for quickly and easily.
One of the key “wins” that Structured Data can provide is more effective searching. Categories don’t effectively describe media and their content the way structured data can.
Search results show media that closely fit what the user searched for (as long as those files actually exist on Commons).
Value / Rationale:
Condition of Satisfaction:
***** SD win (major selling point for Structured Data) *****
[Epic] Multilingual search
[Parent Saga] Effective search
As a user, I want to be able to search in my native language and find many media results in the search results, regardless of what language they were originally described in.
One of the key “wins” that Structured Data can provide is the ability to provide data labels that can be used in multiple languages. This is one of the biggest problems with the current category system in Commons.
Search results show media that were originally assigned metadata in a language other than the user’s default language
Value / Rationale:
Condition of Satisfaction:
***** SD win (major selling point for Structured Data) *****
[Story] Semantic Search results
[Parent Epic] Multilingual search
As a user, I want the meaning of my search terms to matter more than language. If I search “cat”, show me media also labeled with “gato”, “chat”, etc.
Commons is an international project with content labeled in multiple languages. But visual media is often universal and needs no translation. With Structured Data we can expose all media to all users.
Users can search for terms in one language and still receive search results for the same “meaning” but in different languages.
Value / Rationale:
Condition of Satisfaction:
[Epic] Build your own query
[Parent Saga] Effective search
As a user, I want to be able to effectively pick and choose which structured properties and values I want to search for. Ex: “I want videos created after 2004 that depict cats but not dogs.” This is possible with things like Wikidata Query service, but that is too techie. I want an easy to use UI for this.
With the addition of Structured Data, media files on Commons will have a number of new fields associated with them, many of which will be added by Community members over time. It won’t be feasible to account for all in an HTML form.
Users can include practically any statement (and desired value/range for that statement) in their search query
Value / Rationale:
Condition of Satisfaction:
***** SD win (major selling point for Structured Data) *****
[Story] Filter search by file type
[Parent Epic] Build your own query
As a user, I want the ability to apply search terms and requirements to ONLY images, or only videos, etc.
To help users narrow down the search results and find what they want faster by excluding media formats they aren’t interested in.
Users can filter / exclude media formats as part of the search conditions..
Value / Rationale:
Condition of Satisfaction:
[Story] Filter search by “instance of”
[Parent Epic] Build your own query
As a user, I want the ability to apply my search to only photos, or maps, or portraits, or landscapes. Example: I only want photos of Seattle created in 2014, not maps (even though both are images).
To help users narrow down the search results and find what they want faster by excluding “types” or categories of media they aren’t interested in.��This would possibly rely upon/make use of the “instance of” property in Wikidata (P31) or some other similar but better suited property.
Users can filter / exclude “types” or “classifications” of media as part of the search conditions..
Value / Rationale:
Condition of Satisfaction:
[Story] Find/filter images by license/permission
[Parent Saga] Effective search
As a user, I want to be able to search/filter for images by license type (or permission). Ex: show me images of cats that are ONLY Public Domain.
Viewers often have a specific use scenario for the media they’re looking for, and they don’t want to be bothered with media that don’t fit the criteria.
This is currently not possible in Commons but can be done once we have structured licences.
EX: I intend to use this image for commercial purposes and don’t want any images that have ShareAlike licenses.
User has some UI tools that allow them to filter by license type or permission.
Value / Rationale:
Condition of Satisfaction:
[Story] Multilingual auto-suggest
[Parent Epic] Multilingual search
As a user, I want to view auto-suggest text input from structured data in my language so I can see suggestions for terms as I type.
Auto-suggest will help users know that there are many structured concepts (Q items) that may already exist for their search terms.
As user types search terms, or “tags”, they can see a list of existing Wikidata Q items that are used in Commons depicts statements.
Value / Rationale:
Condition of Satisfaction:
***** SD win *****
[Epic] Improve Search results layout
[Parent Saga] Effective search
As a user, the current search result layout (1 column, small images, no preview capability) makes it hard for me to quickly find the image that I want.
The presentation of search results on Commons is problematic and has been a source of user complaints for years (see https://phabricator.wikimedia.org/T104565 as just one example). Improving it will greatly change the Commons user experience for the better.
Users see a search results layout that allows for at some of the following:��- larger thumbs�- less text that is irrelevant to search context�- ability to preview larger version and/or more metadata�- more than one column of images (as long as the device can display it)
Value / Rationale:
Condition of Satisfaction:
[Story] Larger thumbnails/previews
[Parent Epic] Improve Search Results layout
As a user, I want adequately sized previews of images and videos in my search results so I can really see what a file contains before I go the file page.
Currently, search results on Commons include thumbnail images. However:��- at higher resolutions these thumbnails are too small��- more detailed imagery is hard to see
Previews of files in search results allow for larger images.
Value / Rationale:
Condition of Satisfaction:
[Story] Structured Data on Media Viewer
As a user, I want to view structured data in Multimedia Viewer in my language so I can easily view/understand contextual information.
One of the key “wins” that Structured Data can provide is the ability to provide data labels that can be used in multiple languages. This is one of the biggest problems with the current category system in Commons.
Structured data labels/tags appear on the Multimedia Viewer
AND
Structured data “tags” show aliases/labels in the users language whenever possible.
Value / Rationale:
Condition of Satisfaction:
[Story] Related Images
As a user, when viewing an individual image in file page or MediaViewer, I would like to see a short list of thumbnails for related images.
This is a “discovery” feature that highlights other images the user may be interested in. With Structured Data, we can show related images across different categories.
File pages and MediaViewer show a short list of thumbnails for images that are tagged with properties similar to the one the user is currently viewing.
Value / Rationale:
Condition of Satisfaction:
***** SD win *****
Viewer/Consumer Stories
[Saga] Re-use
As a viewer/consumer, I want the ability to embed media with data in the same language as my website, blog, or Wikimedia project.
The data can be just as valuable as the media (sometimes more so). Enabling embed tools that can include the image/video/sound, a gallery, and accompanying data like descriptions and attributions in multiple languages would be very valuable for international audiences.
Provide a tool that allows users to embed media content along with labels, descriptions, attributions, etc. The tool should provide data in the website’s target language (when available).
Value / Rationale:
Condition of Satisfaction:
[Epic] Easy but flexible embed code
[Parent Saga] Re-use
As a viewer who has their own website, I want the ability to select some display options for a Commons file and receive code that allows me to embed media on a web page, with all the display work done for me.
More technical users who have their own HTML sites don’t necessarily want to host the Commons media file themselves, they just want to point to it via code.
Provide a tool that allows users to:�- select options for how they want the media displayed on their site
-- display code generated by the above options.
- preview the end result of the code
- easily copy the code�
Value / Rationale:
Condition of Satisfaction:
[Epic] WordPress plugin for Commons Media
[Parent Saga] Re-use
As a viewer who has a WordPress blog, I want the ability to search for, select, and embed Commons images from within the WordPress admin UI.
Less technical website owners tend to use packaged blogging platforms with WYSIWYG interfaces, WordPress being the most prominent.
To best serve these users, a WordPress plugin that integrates with their current media selection workflow would be optimal.
WordPress users can install a Commons plugin that allows for:�- Search�- Embedding�- Attribution via a caption underneath the media�- mobile friendly views
Value / Rationale:
Condition of Satisfaction:
Editor Stories
[Story] Batch editing
As an editor or curator, I need the ability to edit structured data entries for multiple files at the same time so I don’t have to do them one by one and hate my life.
It’s common for multiple media files to need the same information fixed/changed. It would be very useful to allow editors and curators to change entries for multiple files in a single interaction.
Users have the ability to select multiple files and change/add common data entries in a single action.
Value / Rationale:
Condition of Satisfaction:
[Epic] Find things that need fixing
As an editor or curator, I need the ability to find files that need adjustments to their structured data for quality purposes.
Most editors and curators have particular patterns of organization or quality that they want to patrol, and change. Being able to automatically and semi automatically build worklists, and then systematically work through those worklists, allows for quality improvements at scale.
Create work queues using maintenance tags or queries, that can then be patrolled and reviewed systematically or shared with other community members.
Value / Rationale:
Condition of Satisfaction:
[Story] Data adjustments by editors
As an editor, I need the ability to change structured data entries on images uploaded by all users.
Structured data, like any input from users, could be wildly incorrect, abusive, etc. Editors/admins need to have the ability to change EVERYTHING.
There are no structured data fields Editors and admins can’t change.
Value / Rationale:
Condition of Satisfaction:
Uploader Stories
[Story] Optimized Tagging suggestions
As a Photographer, I want suggestions for how to label or “tag” images so my work can be found easier and more often.
Photographer needs for “tagging” images may be different than what casual uploaders require. Photographers tend to care more about having their images found, used, featured, etc.
Suggestions for data tags that would help the image appear more in search make the Photographer more likely to add extra data.
User is shown some on-screen functionality that displays some sort of search “optimization” tags.
Value / Rationale:
Condition of Satisfaction:
***** SD win *****
[Story] Guided “tagging”
As a casual uploader, I want to be able to upload media and add data to it without having to think too much about how to label/categorize it.
Initially, the data we get for an image will only be as good as the user’s willingness or ability to describe the image with various properties.
Some kind of “suggested” tags functionality makes it easier for the user to tag images, which will in turn increase the amount of data.
User is shown some on-screen functionality that displays some sort of recommendation for tags relevant to this image.
Value / Rationale:
Condition of Satisfaction:
***** SD win *****
[Epic] Attribution
As an uploader, I want viewers/consumers to have tools that allow them to easily and consistently attribute me or my organization when reusing content.
Proper attribution is a key problem media creators and uploaders deal with; they want to be sure the world knows where the image came from when someone uses it on the internet.
The current attribution generator leaves a lot to be desired in this regard. Structured Data presents an opportunity to improve upon this.
Viewers are given a tool that provides proper attribution that fits the given license, via automated code (HTML, WordPress widget, etc.)
Value / Rationale:
Condition of Satisfaction:
***** SD win *****
[Epic] Batch upload
As a Batch Uploader, I need to be able to upload and add structured data to multiple media items while doing batch upload with my favorite batch upload tool (including PattyPan, GLAMPipe, future tools, etc.)
Batch processes are meant to do as much work as possible at one time. Currently categories, templates, and other data items are added via batch upload tools so the user doesn’t have to upload then go back to Commons and add a bunch of data to each individual image.
These tools are generally written by 3rd parties. We would just need to empower them to update their tools to support SD.
Batch Uploaders have functionality that allows them to add structured data properties to each media file via their existing upload tool.
Value / Rationale:
Condition of Satisfaction:
[Epic] Data Download
As a Batch Uploader at a GLAM, I want the ability to download structured data for media in our Commons collection so we can put that info in our own databases.
This is a common request seen in interviews with GLAM users. It is currently difficult to get info back out of Commons, and once structured data is implemented, the crowdsourced data entries that accumulate over time become valuable to the GLAMs and they’d like to download all that data for their own use.
GLAMs (and any other person/group who wants to) has a mechanism for easily downloading files (individually or in batch) along with all their structured data in a format that maintains the structure of the metadata.
Value / Rationale:
Condition of Satisfaction:
[Story] Show data verified by GLAMs
As a Batch Uploader at a GLAM, I want the ability to differentiate between data items that come from my organization and data that was crowdsourced.
GLAM users would like some sort of reputation/verifiability capability for data ("How can we communicate that one field comes from our catalogue, original creator’s knowledge, or a volunteer verified with research, while the other data is from crowd sourcing?")
GLAMs and other reliable data providers have a mechanism for indicating that data for a given media item came directly from that GLAM or other source.
Value / Rationale:
Condition of Satisfaction:
[Epic] Support Commons Android app
As as a user of the Commons Android application, I need to add information about my uploaded files in my own language(s), and have speakers of other languages be able to find and understand my uploads as well.
The Commons Android app allows uploads. Multilingual Structured Data features need to be available for this app to use.
Commons Android app users can, at the very least:��- add and view Multilingual caption/summary fields
-- add and view the multilingual “depicts” properties for the file.
Value / Rationale:
Condition of Satisfaction:
[Saga] Improved file usage metrics
As a frequent batch uploader (including GLAM), I want detailed metrics on how the files I uploaded are being used/viewed/downloaded.
Currently, collecting metrics about item traffic and re-use is done on an ad-hoc basis: there is no way to automatically record usage data at specific intervals and track (or visualize) activity on an ongoing basis.
User is provided some functionality to view and track media file usage.
Value / Rationale:
Condition of Satisfaction:
***** SD win *****
[Epic] Usage data dashboard
[Parent Saga] Improved file usage metrics
As a (GLAM) contributor to Wikimedia Commons, I want a regularly updated reporting dashboard, without having to generate reports from irregularly updated tools.
Some GLAM institutions expect to be able to view "status reports" of the impact their donation has had at regular intervals. Currently, these reports must be manually assembled
Availability of a dashboard for monitoring usage of files a user has uploaded.
Value / Rationale:
Condition of Satisfaction:
Curator Stories
[Epic] Add structured data in multiple languages
As a curator, I want to be able to add structured data for media on Commons in languages I am fluent in so that others who speak those languages can find and understand content.
Curators who are looking to organize and improve the data on each media item will want to contribute to one of the biggest ongoing efforts on Commons – adding/editing relevant information in multiple languages.
For Structured Data, this would apply to data values like Description, label aliases, etc.
Curators (and other users) have the ability to add additional text in multiple languages to new or existing Structured Data entries where relevant (regardless of what language the original media metadata was written in).
Additionally, users who are fluent in multiple languages can add info in all those languages at once.
Value / Rationale:
Condition of Satisfaction:
[Epic] Convert existing data to structured data
As a curator, I need tools that will help in converting current unstructured or semi-structured data to structured data in file pages on Commons.
A lot of information is contained in data entities that are unstructured. Categories are one example. Some categories are connected to Wikidata, but most aren’t. Many users have favorite workflows that involve manually gleaning info out of categories. This can be much improved by converting this data to Structured Data either within Wikidata or MediaInfo.
Curators (and other users) have the ability to convert existing data in wikitext to Structured Data.
This ability can be provided by Commons itself or volunteer tools.
Value / Rationale:
Condition of Satisfaction:
[Epic] Monitor Changes to files
As a curator or GLAM user, I want to have a publically available feed/watchlist that is both human and machine monitor-able to watch and review the changes to a batch of files (or just one).
This would allow GLAMs or project/event organizers to keep an eye on “collections” they manage.
Question: should other users be able to see what files a given user is watching? If so, should there be an option to make the list of watched files private?
GLAMs (and any other person/group who wants to) have access to a feed that lists changes to each file in a collection.
Value / Rationale:
Condition of Satisfaction:
Toolbuilder Stories
[Epic] API support for Structured Data
As a tool builder, I need to have APIs that allow me to use my software to add and/or edit structured data for media on Commons.
Commons functionality is heavily supplemented by software from volunteer developers. One of the goals of the Structured Data project is to empower these developers to add useful new functionality to their existing projects and also build exciting new tools that extend the functionality of commons.
Developers have a stable and well-documented API that exposes all aspects of the Structured Data implementation on Commons.
Value / Rationale:
Condition of Satisfaction:
Misc. stories
(important but not directly related to SDoC)
[Epic] Improved documentation for devs
As a tool builder, I need to have documentation that covers the unique technological and community aspects of developing tools based on Commons.
The barrier to entry for some developer partners is often high, in part because the platform documentation doesn't account for both the technological and social dynamics involved with Wikimedia projects.
Developers have thorough documentation that exposes all aspects of developing for Commons.
Value / Rationale:
Condition of Satisfaction:
[Epic] Curated Pages
As a curator or GLAM user, I want to build curated content pages for consumers that I can design using both semi-automated (data-based) tools and manual editing.
.
One of the common activities on Commons is building things like Category pages and Gallery pages. These pages offer hand-curated interpretations of what exists on Commons. Organizations like GLAMs may also want to create "Institutional Galleries/Landing pages" that act as small exhibits of what they have on Commons.
Users can create “Landing Pages” for groups of content on Commons using a mix of suggestions/templates from data and manual editing.
Value / Rationale:
Condition of Satisfaction:
Notes and Questions
- No stories specifically for Admins. Did I miss something or is this not a problem?
- Similarly, do we have more wins that directly apply to Curators? Things that are a “wow” for them?
- There aren’t many use cases for developers. Is there anything that Structured Data makes easier for them, or something new that wasn’t possible before?