Seminar: Norman Su

Pushing the Boundaries of "Doing" Research Papers in Computing

In computing, papers (conference papers or journal articles) are the primary means by which researchers disseminate new knowledge. Despite greater acceptance of alternative mediums to communicate research (e.g., blogs, podcasts) to a wider audience, academics are primarily judged by their peer-reviewed publications. We are incentivized to do innovative research but less so on innovative ways to communicate research. Lest we face the wrath of Reviewer 2, it is often safer to draw on established norms of writing to communicate research.

In this talk, I describe a series of studies with colleagues that scrutinize how we 'do' papers in two fields of computing: HCI and Computer Vision. I will also describe efforts to push the boundaries of what can be written, within the constraints of the (usually PDF) paper. Lastly, I will close on future directions for writing papers that have a more active and affective impact on their readers. (edited)