The Lost Visions project aims to make more than a million images from the British Library’s collection available for the public to search online. Working with a dataset consisting of 68,000 digitised volumes (approximately 25 million pages) brings with it many of the challenges inherent in understanding and
contextualising ‘big data’. One of the ways in which the project aims to enhance the available metadata is through the trial of crowdsourcing methods for assigning tags.
The meeting raised a number of important questions about how online communities might be encouraged to participate in the process of generating iconographic tags for the images in the collection, and also within the digital humanities more generally.
Representatives from Metadata Games (http://www.metadatagames.org/), an online gaming platform for gathering data for digital media archives, provided an overview of the work they have been doing to encourage players to visit archives and explore content while contributing to existing data. In addition to crowdsourcing games, we also considered ways of enhancing the interest in the British Library’s collection of images which has already been expressed via social platforms such as Flickr and Wikimedia Commons.
The crowdsourcing workshop will help to engage participants with the project through a series of interactive and
collaborative tagging exercises. However, one of the challenges faced by engagement in the arts and humanities is that care must be taken to avoid limiting engagement to physical attendance at events: the workshop will function as a platform for encouraging and maintaining future participation in the project within the online community rather than as a one-off occurrence.
In addition, discussions about the theme for the workshop raised still more issues about the relationship between tagging as a
means of public engagement and tagging as a means of refining and enhancing existing data in order to maximise accessibility for potential users. The Lost Visions project aims to make thousands of neglected images more widely visible and it is important that crowdsourcing contributes to the output of a searchable database with accompanying metadata as well as to the increase of public engagement. As such, the workshop will provide a valuable means of evaluating the practical problems of user-generated tagging: it is hoped that participants will be able to offer feedback on the webpage interface and tagging process as well as some of the wider issues relating to motivation, recognition and community in crowdsourcing in the digital humanities.
The meeting provided us with plenty of ideas and, in some ways, raised as many questions as it answered (not least about how to decide on an appropriate theme for the workshop that will both engage participants and provide us with relevant feedback, but more on that later…). What is clear, at least, is the potential of crowdsourcing for digital image archives, not only in terms of refining and contributing to existing data but as a means of deepening public engagement with visual culture.