Grants:IEG/Improve 'Upload to Commons' Android App

From Meta, a Wikimedia project coordination wiki
Jump to: navigation, search
statusselected
Improve 'Upload to Commons' Android App
IEG key blue.png
summaryImprove and promote Upload to Commons Android app to facilitate contribution of useful images to Commons.
targetCommons, Wikipedia, Wikivoyage
strategic priorityincreasing participation
amount7800 USD
granteeMisaochan
advisorSyced
contact• josephinelim86@gmail.com
this project needs...
volunteer
join
endorse
created on07:45, 2 March 2016 (UTC)
round 1 2016

Project idea[edit]

What is the problem you're trying to solve?[edit]

Mobile phone cameras are the most commonly-used type of camera currently, so there is a rising need for an app that allows Commons contributors to easily submit their photos directly from their phone to Commons without needing to go through the web interface. The Upload to Commons app (Android) attempts to fulfill this need.

Over the time I have contributed to this app (Oct 2015 - present), I have seen the number of active installs increase from <20 to >170. In my opinion there is a growing interest in this app, and it encourages people to make more contributions to Commons. Currently there is an average of 50 pictures a day being uploaded to Commons using this app, which cover a wide variety of subjects from around the globe (e.g. in the last 3 days I have seen pictures from Jakarta, Nevada, Madrid, Bahraich, Bamberg, etc).

All of this is great - but the app has potential to do even better. There are a few long-standing bugs that need fixing, old crash reports that could be looked into, and there are several enhancements that would very much improve the app's usability and increase its usefulness. Also, some publicity is needed to inform non-Wikimedians about this app, as it is one of the easiest and least-intimidating ways for a WM newcomer to start contributing.

What is your solution?[edit]

We plan to:

  • Fix long-standing bugs and look into old crash reports to try and prevent them from recurring, to provide a smoother user experience and avoid frustrating users
  • Make several enhancements to the app, including suggestions by the current app maintainer and users. These enhancements involve making the app more user-friendly and newbie-friendly, new location-based features, reducing the workload for Commons volunteers, and further enhancing categorization.
  • Increase awareness of non-Wikimedians about this app, as it provides a nice easy way for newcomers to start contributing to WM and Commons.

Project goals[edit]

  • Make this app more user-friendly and convenient to use for both newcomers and experienced users
  • Encourage submission of appropriately-categorized, unique and useful pictures to the Commons database
  • Grow the contributor base by promoting this app as an easy and non-intimidating way for newcomers to start contributing to WM and Commons

Project plan[edit]

26 weeks @ 20 hours/week (50% commitment). All of the tasks below are associated with their corresponding GitHub issue #.

Phase 1 : Fix existing bugs and crashes - (4 wks)[edit]

There are several long-standing bugs recorded in our GitHub issues page, such as:

  • Empty description/license/category fields in newly-uploaded pics (#12)
  • Newly-uploaded pic replacing some of the other pics in My Uploads screen (#25)
  • Max 500 files scanned (#54)
  • "Template: According to EXIF data" analyzing EXIF data incorrectly (#72)

I plan to fix these bugs and look into old crash reports (on the commons-app-android Google Groups forum) to try and prevent them from recurring.

Phase 2 : Improvements/enhancements - (18 wks)[edit]

2(a) Make app more newbie-friendly - (5.5 wks)[edit]

  • Allow account creation from within the app itself (#48). Users must be logged in to their Commons (or global) accounts to upload images, which is a necessary precaution but does create a barrier to entry. A 'Sign up' link to the web registration form was added in a recent patch, but that redirects users to their browser and is less convenient. To reduce this barrier to entry, we intend to embed account registration in the app to make account registration as fast and straightforward as possible.
  • Warn users if they submit image without selecting categories (#61). Users submitting images without categories is an issue, as that makes it difficult for the pictures to be used appropriately. New users especially might not realize how important categorizing their pictures appropriately is (I sure didn't when I first started!). We plan to inform users of this by displaying a warning if they submit an image without selecting any categories, informing them that their images are much less likely to be usable and asking if they want to go ahead with the submission.
  • Warn users if they press back button without selecting categories (#58). Pressing the back button on the category screen by mistake will irreversibly take the user back to the main screen without allowing them to select categories, whereas the upload has already begun. To prevent this from happening, we plan to also display a warning if the user does that, asking them if they are sure they want to go back.
  • Add page to tutorial to educate new users on what types of images are useful (#95). New users might be unaware of rules regarding copyright violations and selfies, and they might upload pictures violating these rules that require deletion. This could be reduced if we educate new users on (1) what types of images are useful, and (2) what types of images will certainly be deleted. We intend to add a page to the tutorial (that is displayed when a new user installs the app) for this purpose, Also, we are hoping to collaborate with folks who are experienced in educating new users, in order to maximize the effectiveness of the tutorial. (Edit 30 May: Potential collaboration with Pine in August)

2(b) New location-based features - (5 wks)[edit]

  • Display nearby wiki articles lacking photos (#73). We plan to display a list of nearby places (wiki articles) which lack photos, so as to inspire users to upload needed images. We also plan to sort the places without a photo by number of page views / month, to encourage photos of more important places. The list of nearby places lacking photos will be obtained via either the wikishootme API or the Wiki Needs Pictures API, and we plan to create a navigation drawer to house the link to this list (along with links to Settings and Feedback).
  • Add geocoding template (#35). The web-based Upload Wizard automatically adds the Geocoding template to uploaded images, but the app doesn't. The newly-implemented GPS categorization in the app depends on finding nearby pictures (i.e. pictures with similar coordinates in the Geocoding template) in order to suggest categories, so not having this template hampers categorization of future pictures. We plan to automatically add Commons Geocoding templates to images from their EXIF location data.

2(c) Enhance titles and categorization - (6 wks)[edit]

  • Option to copy title and description from previous upload (#65). Users often upload more than one image for the same topic, usually consecutively. This could make it easier for users to do so without having to retype everything. We intend to implement this via saving the previous title/desc and having a button on the Title/Desc screen that says 'Use previous title/desc'.
  • Suggest categories based on entered title (#43). People usually describe their image pretty well in the Title field, so it would be useful to perform a category search on the content of the Title field. The categories obtained could be shown as suggestions below the nearby categories, which could potentially save the user a lot of time in searching for appropriate categories. For this purpose, we intend to reuse the API calls from the recently-implemented flexible category search.
  • Filter out year categories (#47). Sometimes, suggested categories include "X in 2006". Those are very unlikely to be appropriate, so it would be best to filter them out. This would be done by filtering all categories that contain a 4-digits number that starts with "19" or "20".
  • Headers to distinguish where suggestions come from (#76). The category screen already has suggestions for recently-selected categories and nearby categories, and we plan to add 'guessed from title' categories in this project. Chances are that a user already knows what type of suggestion will be correct. e.g. if the last upload was of the same thing, they would be looking for a 'recent' category, and if they are in a famous place but photographing something unrelated to the place (e.g. a vehicle) then they can ignore all 'nearby' suggestions. Having sections will thus make selection faster.

2(d) Reduce workload for Commons volunteers - (1.5 wks)[edit]

  • Prevent uploading multiple copies of exact same image (#77). It was suggested in the Commons Village Pump that there is an issue with contributors (presumably unintentionally) uploading multiple duplicates of the same image, and it was requested that we implement a check to see if the exact same image already exists in the Commons database. We intend to implement this suggestion by checking for similar SHA1 hashcodes using this API. If a match is found, that means the picture is a duplicate, and the user will be alerted of this.

Phase 3: Increase publicity - (4 wks)[edit]

After the app has been nicely polished up, we intend to increase non-Wikimedian awareness of this app. This will be done by:

  • Creating a basic landing page for the app (most likely on GitHub, so future contributors can edit it while having more flexible styling than a wiki page), with a simple tutorial on how to use it and what the submitted pictures are used for
  • Creating a Youtube video demonstrating the use of the app. This video could also be embedded in Play Store. (#86)
  • Contacting several bloggers, app review websites, and tech-related magazines (e.g. techcrunch, lifehac.kr etc) to ask if they would be interested in writing about the app

Budget[edit]

Developer salary: 15 USD/hr

  • Phase 1: 4 wks * 20 hrs/wk = 80 hrs = 1200 USD
  • Phase 2: 18 wks * 20 hrs/wk = 360 hrs = 5400 USD
  • Phase 3: 4 wks * 20 hrs/wk = 80 hrs = 1200 USD

Total = 1200 USD + 5400 USD + 1200 USD = 7800 USD

Community engagement[edit]

I have been communicating with users of the app in the Google group forums as well as on the GitHub issues page and Phabricator. Some of the enhancements and bugfixes in this proposal were initially suggested by users of the app.

In addition to the main tasks listed, there will be a regular cycle of obtaining feedback from users and solving any problems they might have with the new implementations. After each phase is complete, I plan to announce the release via the wikitech-l, commons-l, wikivoyage-l and mobile-l mailing lists, to request for volunteers to test the new features. Crash/bug reports and feedback will be welcome as usual on the Google group forums or our GitHub page.

Sustainability[edit]

This project aims to make it easier for people to upload images to Commons, and to recruit and educate new Commons contributors - the benefits of this should continue long after the project is completed. Also, the project code will be available on an open source Apache license on GitHub, and will be well-documented so that future code contributors can be brought up to speed.

Measures of success[edit]

  • All tasks completed
  • 50% increase in active installs between the start and end of the grant period (tracked via Google Play Developer statistics) -- Note: 217 active installs at start of grant period (20th June)
  • 80+ pictures being uploaded daily through the app (tracked via CatScan, currently hovering around 50/day) that are of acceptable quality (not reverted after 3 days)
  • 10+ new contributors who have never contributed to Commons before. (Optional - needs to be measurable on WMF side due to potential privacy concerns re: collecting this data on the app)
  • Less than 10% of pictures uploaded by new users (who have never contributed to Wikimedia before) are unusable (selfies, copyvios, etc). This will be measured over a period of 2 weeks at the end of the grant.

Get involved[edit]

Participants[edit]

Grantee - Josephine Lim (User: Misaochan, GitHub: misaochan) - I'm a Computer Science graduate from the University of Auckland, New Zealand. I have worked on this app since Oct '15, initially as a part-time contributor and then full-time from Dec '15 - Mar '16 as an Outreachy intern, where I successfully completed my project under the mentorship of Stephen Niedzielski and Nicolas Raoul. I have implemented several new features and bugfixes that are included in the current version of the app.

Advisor - Nicolas Raoul (User: Syced, GitHub: nicolas-raoul) - Current maintainer of the Upload to Commons Android app and Wikimedia contributor since 2005.

Campaigns[edit]

We have received feedback that it might be useful to implement campaigns in the app. We are open to the prospect - however, a campaigns feature was in the app previously and was removed because it was unused. This is being discussed on GitHub, and we welcome your input. If possible, we hope to be able to talk to someone who is familiar with campaigns, and who is willing to test our old implementation of campaigns and provide feedback on whether it should be reimplemented.

Community Notification[edit]

Endorsements[edit]

Do you think this project should be selected for an Individual Engagement Grant? Please add your name and rationale for endorsing this project in the list below. (Other constructive feedback is welcome on the talk page of this proposal).

  • Community member: add your name and rationale here.
  • In today's world, most of us use our phones to click images of certain entities that require representation on the Commons. I use my Android phone for photos of buses, Toll plazas, etc. This app would make things a smoother process for contributors. --Rsrikanth05 (talk) 05:16, 4 March 2016 (UTC)
  • Bug fixes and enhancements needed! Riley Huntley (talk) 03:12, 7 March 2016 (UTC)
  • I'm trying to do my best with other fellow wikimedians to "fix" the lack of pictures, but this is a well-known other side of the problem. And it's time to fix it as well.−−Alexmar983 (talk) 08:37, 18 March 2016 (UTC) User:Misaochan I will notify other communities as I did with my grant, fell free to contact me if you want some help with translations in other languages of your work.−−Alexmar983 (talk) 08:39, 18 March 2016 (UTC)
  • WMF can increase the number of uploads and general participation by showing articles nearby which lack photos. I think this proposal will address the goals of increasing photo coverage on articles and increasing general editor participation. Justin Ormont (talk) 00:56, 7 April 2016 (UTC)
  • This app really needs to be improved. Feldo (talk) 15:02, 18 April 2016 (UTC)
  • I found this app entirely off-wiki, on Google Play. It has potential and, above all, we need an app like that. During wiki workshops we usually ask attendees to browse their phones in order to find a pic or two they could share. Some of them are truly amazing - and could fill in the gaps we have perfectly. Take this origami snake for instance. Some lady had it on her phone, she did it with her kids once. It turns out we've had the entire manual on Commons - but not a single picture of the finished snake. Here's some examples from the latest wikiworkshop we organised: 1, 2, 3, 4... People have such gems on their mobile phones; if only there was a way to allow them to share those pics more easily... Halibutt (talk) 23:02, 10 May 2016 (UTC)
  • I completely support this project based on recent experience - I would add two other parts to it if they are feasible - user registration via app and easier localization and forking for language versions - it seems that localizing apps might be easier than getting (esp. weaker) communities to fix language specific instructions, templates, license details etc. on Commons. Shyamal (talk) 13:38, 18 June 2016 (UTC)
  • Strong Support. Would be good if users could set up a Radius where they get push notifications if a missing image location is for example just 1 km away. --Tobias1984 (talk) 09:40, 3 July 2016 (UTC)