03 March, 2016

Visualizing Ekphrasis

One of the most rewarding moments as an administrator is when your colleagues (here: the wonderful Carl DiSalvo, Yanni Loukissas, and Lauren Klein) invite you to participate in a workshop during which you can expand your current horizons. The workshop encourages leading humanities scholars with visualization designers and researchers to imagine new forms and platforms capable of portraying the humanistic dimensions of culturally significant data-- an increasingly important challenge in the fields of digital humanities and information visualization. The Humanities Data Visualization workshop is funded by Georgia Tech's GVU Center, the Institute for People and Technology, the School of Literature, Media, and Communication, the Ivan Allen College of Liberal Arts, and the Intel Center for Social Computing (ISTC-Social). And I find myself in a workshop on Revising Ekphrasis: Ekphrastic poetry, or poetry that describes visual art, present multiple layers of meaning within (and beyond) the bounds of the poem. This dataset contains the text of 4,500 poems filtered through a topic model (a technique derived from the field of machine learning). While the metaphorical language of poetry resists the thematic analysis that would usually follow, the model’s underlying statistics open up new opportunities for the analysis of ekphrasis in a wider range of poems. Team members: Lisa Rhody (data), Paolo Ciuccarelli, Alex Endert, Gabby Resch, Jen Sterling, Richard Utz, and Zixuan Wang.  Project website: