Volunteer classifications of images from the Cropland Capture game

Each entry represents a single classification of a single image by a volunteer rater.

The dataset contains six columns:

imgid:     The unique identifier for each image used in the Cropland Capture campaign
userid:    The unique identifier for each volunteer in the Cropland Capture campaign
rating:    The answer provided; can be only one of the following:
         1: yes cropland
         2: no cropland
         0: maybe 
date:        Timestamp of the rating
ratingid:    The unique identifier of the rating (this is different for each data row)
platform:    What interface did the volunteer use to provide this rating?
         1: iPhone5
         2: iPhone, other models
         3: iPad
         4: Browser
          >100: Android; different numbers indicate the screen size in pixels 

For more information, please see the following publications:

Salk, CF, T Sturn, L See, S Fritz (2017).  Limitations of majority agreement in crowdsourced image interpretation. Transactions in GIS, 21: 207–223.

Salk, CF, T Sturn, L See, S  Fritz (2016).  Local knowledge and professional background have a minimal impact on volunteer citizen science performance in a land-cover classification task.  Remote Sensing, 8: 744.

Salk, CF, T Sturn, L See, S Fritz and C Perger (2016).  Assessing quality of volunteer crowdsourcing contributions: Lessons from the Cropland Capture game.  International Journal of Digital Earth, 9(4): 410-426.