Parsing Related Tags Networks from Flickr® to Explore Crowd-Sourced Keyword Associations

Abstract

With the broad popularization of content-sharing social media platforms, researchers have developed methods to extract information from social related metadata. One approach is to extract related tags networks from freeform “folk” tags (keywords) used to describe imagery from Flickr®. These tag networks, graphed at 1, 1.5, and 2 degrees, show co-occurrence of related tags (above a certain threshold) to a target seeding tag. This work describes how these networks may be acquired and different ways that the extracted data may be used analytically in a digital humanities context.