演讲者： Professor Tim Baldwin (The University of Melbourne)
地点： Room 309, ICST Building
Two of the key tenets of modern-day democracy are the active participation of the people in politics and civic life, and political pledges to the people by political parties, e.g. in the form of manifestos. In this talk, I will present recent work on two political text regression problems: (1) prediction of the right-left leaning of a political party on the basis of their manifesto document; and (2) prediction of the popularity of an online petition, in terms of the number of votes it will receive. In both cases, we propose neural models with mixed objective functions incorporating auxiliary tasks, and in the case of manifesto analysis, further smooth the model output based on
contextual and temporal constraints using probabilistic soft logic.
Tim Baldwin is a Professor in the School of Computing and Information Systems, The University of Melbourne, and Associate Dean (Research Training) in the Melbourne School of Engineering. He is also Director of the ARC Centre in Cognitive Computing for Medical Technologies, in partnership with IBM Research. His primary research focus is on natural language processing (NLP), including deep learning, social media analytics, computational lexical
semantics, and topic modelling. Tim completed a BSc(CS/Maths) and BA(Linguistics/Japanese) at The University of Melbourne in 1995, and an MEng(CS) and PhD(CS) at the Tokyo Institute of Technology in 1998 and 2001, respectively. Prior to joining The University of Melbourne in 2004, he was a Senior Research Engineer at the Center for the Study of Language and Information, Stanford University (2001-2004). He is the author of around 350 peer-reviewed publications across diverse topics in NLP, and his research has been recognised with a number of best paper awards, an ARC Future Fellowship (2013-2016), and an IBM Faculty Award.