Night-Knights-dataset

Night Knight is a GWAP that aims to classify images taken from the International Space Station (ISS) into a set predefined categories.

The six possible categories are mapped in the following codes:

• 1 = black
• 2 = city
• 3 = stars
• 4 = aurora
• 5 = iss
• 6 = none

Below we describe the structure of the datasets stored in this repository, detailing the meaning of all the columns of the .csv files.

Dataset1

File with the log of the users' classifications.

Columns description:

• idResource: ID of the image classified
• idUser: ID of the user
• chosenCategory: class selected by the user
• incentive: with possible values "EXTRINSIC" and "INTRINSIC", indicats the type of incentive available when the image has been played. "EXTRINSIC" indicate a period with a tangible reward (i.e. during a competition), "INTRINSIC" indicate an intangible reward like fun.

Dataset2

List of all the images classified with the final classification assigned after having run the aggregation algorithm on all the users' contributions.

Columns description:

• idResource: ID of the image classified
• aggregateCategory: final classification

Dataset3

List of all the images played in the game with the URL to the NASA website.

Columns description:

• idResource: ID of the image classified
• refUrl: url of the NASA page with the image.