What is the Social Media Engine?
The social media engine is a framework that captures some of the conversations about scholarship that are happening in social media and brings some context that can then be used to reorganize a document collection. In our specific case, we are more concerned with the conversations that have ties to the Érudit corpus. The mechanisms that we use are very extensible, so we can easily apply our methods to a different corpus.
What needs is the Social Media Engine addressing?
We have found that the methods for representing documents and disseminating knowledge are changing. Consequently, social media is slowly transforming the scholarly communication system. The frameworks for document collections need to be more dynamic. In this sense, our Social Media Engine assimilates the discussions that are happening within social media into the workflow of a digital collection.
Who are the main target users of the Social Media Engine? Could you give a usage example?
The target users for our system are scholars who carry out and base their research on digital document collections. Currently the access methods in digital libraries are generally focused on the characteristics of the collection rather than including the dynamically changing characteristics of the environment – which we extract from social media. We then use these changing characteristics to reorganize the contents of the collection to respond more adequately to information queries. For example, looking for papers on Canadian literature while discussions over this topic are happening on Twitter (#CanadianLiterature) will yield a different set of results than when trending topics are more concerned with international affairs.
On what concepts and methods is the Social Media Engine based?
The fundamental concepts that the Social Media Engine is based on can be explained using three points. First, our framework yields a list of topics related to individual entries and articles in the corpus by applying textual analysis techniques and topic modeling. Second, our engine connects readers and publications by monitoring social media for trending topics and returning links to open access publications that can be used to feed and enrich the discussion. Finally, our engine identifies trending papers on social media by looking at the number of times that papers on social science topics are shared, saved, liked, or commented on.
What is the work plan for the next steps of the developments?
At this point we have a working prototype that is deployed on virtual machines in Compute Canada’s server infrastructure. Our current efforts are focused on planning rounds of user studies that will ultimately result in improvements of the user interface. We expect to launch the prototype before the end of the year.
What is the best way to stay informed about the project?
The CO.SHS newsletter of course!