Speaker of Workshop 2
Will talk about: Automated annotation and meta-analysis of the fMRI literature: promises and challenges
Tal Yarkoni is a research associate in the Institute of Cognitive Science at the University of Colorado Boulder. His research focuses on developing new methods for large-scale acquisition, organization, and synthesis of psychological and neuroimaging data.
This talk reviews recent efforts to represent and synthesize the published fMRI literature on a large scale. I begin by discussing the pros and cons of existing methods in this area, ranging from fully automated "bag-of-words" approaches to more effortful manual approaches. Particular attention is focused on the fundamental trade-off between data quality and data quantity. I then review a number of applications demonstrating the utility of large-scale synthesis-oriented approaches in helping to map the neural substrates of cognition. Finally, I conclude with a discussion of several open challenges for data mining effort in this area, including better extraction of key metadata from published articles, development of empirically validated cognitive ontologies, and successful hybridization of automated and manual annotation efforts.