Crowdpee

Short term goal: develop a platform that asks users to supply information they are geographically well-suited to provide.

Medium-term goal: improve data coverage of unisex/accessible, etc. toilets in Zurich.

Long-term goal: open the platform to allow other organisations to submit requests for geo-crowdsourcing.

Components:

  • List of points of interest (eg restaurants where we don't know what the toilets are like)
  • Web-based system for data entry and cross-verification
  • Twitter bot (and other bots, eg flickr watchers) to ask people to fill in data
  • Crawler/scraper to find out additional information (contact info, etc.)

Data points of interest:

  • unisex
  • accessible
  • baby changing facilities
  • baby changing facilities for men

Progress

* Started developing the Twitter bot, which currently prints out tweets that happen within about 500m of a point of interest.

* The twitter bot now has an account: CrowdPee. It can identify tweets sent from within 100m of a point of interest and determine which point of interest is closest. It can reply to these tweets with this message: “Hi! We're looking for people to gather info about bathrooms in Zürich. (url of questionnaire)”

* The general website for questionnaires has been set up at nearbysources.com, using Django. Each campaign on the website contains a questionnaire, a tweet, and a thank you message, all of which can be stored in an arbitrary number of languages. The campaign also contains locations of interest, including name, lat and lng data.

* Questionnaire pages have been designed to be simple and easy to complete on both computers and phones.

* The CrowdPee campaign has been set up on nearbysources.com.

* To-do items after discussion at Hack Night 3:

  • Make nearbysources.com questionnaire searchable to add data - done!
  • Enable tweeting at the bot to receive a questionnaire
  • Improve the results display and export function

Data

Team

  • Rachel Knowler
  • Vincent van der Lubbe
  • David Stark