Cornell University
more options

Drew Margolin, Wednesday, November 30, 2016

Wednesday, November 30, 2016 - 4:00pm
Gates Hall 114

Cornell Communication Assistant Professor and Geri Gay Faculty Fellow Drew Margolin will be the Information Science Colloquium speaker at 4 p.m. Wednesday, November 30, in Gates Hall 114. Margolin's areas of expertise include computational social science, social networks, and text mining. From his faculty page: "The internet, and social media in particular, have made individual and institutional discourse visible like never before. Yet the mechanisms that shape the production of discourse — what leads individuals or institutions to speak up, whom do they address, what do they say — is not yet well understood. My research focuses on understanding these dynamics through the quantitative aggregation of collective communication behavior. In particular, my approach emphasizes the role that accountability, credibility, and legitimacy within social networks and communities play in shaping observable discourse."

Title: Satisficing Semantic Search: A Theoretical Approach to Analyzing Text in Networks

Abstract: Digital archives from social media and related sites provide access to communication behavior in a novel form.  In particular, these data often contain both social and temporal relations between statements, enabling researchers to identify, for example, which pairs of individuals tend to speak similarly or which tend to lead and which tend to follow.  This talk will present a theoretical framework for utilizing these observations called satisficing semantic search.   This framework focuses on two key concepts from March and Simon's satisficing search theory as applied to organizations: availability–the extent to which textual material can be found with ease; and aspiration–the extent to which individuals are motivated to find material that precisely represents their underlying views.  After defining and providing examples of these concepts in my empirical research I will argue they can be used to make meaningful inferences about things such as the stability of public opinion, media bias, and the virality of specific messages.