UserSubmittedtags_from_Dec2013_to_March2016.zip (9.11 MB)
BL Flickr image dataset: User Submitted Tags (til March 2016)
In Dec 2013, the British Library Labs project uploaded over 1 million undescribed images onto Flickr. These illustrations, diagrams and decorations were extracted from 65,000 volumes of digitised works.
The dataset is a table of the descriptive and other tags added by contributors to the images in order to better describe them, and contain all the user-submitted ones to date (March 2016).
The data table (in TSV format) has the following columns:
flickrid - The id of the image.
enteredtext - The text typed in by a contributor.
from - the epoch time (in seconds) since the last data harvest before the 'to' harvest.
to - the epoch time of the metadata harvest that the change was detected in
tagid - The tag identifier supplied by Flickr
author - The user's account who added the tag
tag - The simplified version of the text, used in flickr's URLs and so on.
mode - either 'add' or 'del', showing that a tag has either been added since last harvest, or that it has been removed since then.