Project Arclight began with an idea. If researchers can use Twitter analytics to study trends in discussions of contemporary media, then what if we treated historic trade papers and fan magazines like a giant Twitter stream and explored trends in film and media history? We refined this idea, published a paper on it in IEEE Big Data, and received a Digging into Data grant to develop a web app and try to better integrate computational research methods into the fields of film and media history. You can learn more about Project Arclight and read our blog at our website.
Arclight searches the collections of the Media History Digital Library (MHDL), which as of August 2015, has digitized nearly 2 million pages of books and magazines related to film, broadcasting, and recorded sound. The MHDL's collections primarily encompass out-of-copyright works. For this reason, the collections largely cut off after the year of 1964.
In the default view, Arclight graphs the number of digitized pages that a search term appears in a given year. In the normalized view (marked by the percentage icon), Arclight graphs the percentage of pages in which a search term appears based on the total number of digitized pages from that year. We encourage you to explore both of these line graph views, as well as the stacked bar graph.
On the Project Arclight blog, we have described our technical method and interpretive framework for what we call Scaled Entity Search (SES). We encourage you to read both of these posts and our forthcoming journal articles that utilize Scaled Entity Search to explore the histories of American radio and early cinema. But if you'd like to get started quickly, we would encourage you to think about the relationship between your search term and the digitized books and magazines you are searching within. To make this process easier, Arclight allows you to click on any point in the graph and open the matching page results within the MHDL's search platform, Lantern.
Arclight Software Development Team
Project Directors:
Charles Acland and Eric Hoyt
Interface Design and Programming:
Kevin Ponto and Alex Peer
Search Index Development:
Eric Hoyt, Kit Hughes, Derek Long, Peter Sengstock, Tony Tran
Thank you also to broader team and
Finally, we wish to thank IMLS, SSHRC, the University of Wisconsin-Madison Office of the Vice Chancellor for Research and Graduate Education, Concordia University Media History Lab, and Media History Digital Library for their support in making this project possible.