Two Years on—Developing Metrics for Crowdsourcing with Digital Collections

Abstract

The National Library of Finland has been supporting crowdsourcing from the spring of 2014, when a new version of the presentation system of digitized newspapers, newspapers and ephemera was released at http://digi.kansalliskirjasto.fi. Since then, there has been steady increase in the usage of the crowdsourcing features. These new functionalities enable any registered user to collect clippings from articles, images and anything else from the digital collections, creating their own set of interesting material. The purpose of this research is to follow how the usage of the crowdsourcing features has evolved over the 2 years when it has been available. In addition, we evaluate how the contractually opened in-copyright materials have been used in crowdsourcing. What the metrics tell, is that there has been steady increase in the interest to the crowdsourcing, based on the metrics of collected clippings. Overall, it is possible to see steady increase over 6-, 12- and 24-month periods. However, there is still significant variance between users—the top users cover 46% of all clippings, whereas there is a long tail of users with just one clipping. Based on this data and the work ongoing in development projects, we feel that crowdsourcing is a viable way for a digital library to attract new kinds of users—both the general public and researchers. However, crowdsourcing requires advocacy, and we should put additional focus on communication, functionalities and availability of the new content.