Speaker: Garron Hillaire

Abstract:

If emotion and subjectivity are the same (or even just closely related), then measurement methods that focus on removing subjectivity may in fact be removing emotion from the measures. In recent work, a measure was built that attempts to harnesses the subjectivity of the crowd through aggregation. The result of the measure demonstrated improvement over a comparable state of the art measure when detecting the complex emotion expression of mixed emotion. There was also a surprising result in removing a significant bias towards neutral detection that was found in the state of the art technology. It may be that research methods that attempt to remove bias are in fact injecting a bias towards neutral detection and sorely inadequate at detecting complex emotion. Perhaps through embracing the subjectivity of people we can develop more effective measures of emotion. With improved measures we will be in a better position to understand the role of emotion in complex processes such as learning.