James Evans, PI
Goals: In his recent work on molecular biology,
James Evans, a faculty member at the University
of Chicago Department of Sociology, used
natural-language-processing tools to summarize
the scientific literature and to assess the ways in
which collaboration shaped the productivity,
persistence, novelty, sharing, and applicationand
commercial-orientation of research results.
Significance: Evans uses CI resources to study the social structure of scientific inquiry. Evans developed a method to parallelize the largescale, long-running queries (written in the Structured Query Language, SQL) of graphstructured citation databases. The parallel method was run at the Computation Institute on the TeraPort facility.
Accomplishments: This method has shown up to a 100X speedup, turning months of analysis time into a few days or less, and permitting significantly deeper analysis and faster verification of results.