Abstract:In this talk, I will present a novel search-and-learning framework for unsupervised text generation. We define a heuristic scoring function that (roughly) estimates the quality of a
In this talk, I will present a novel search-and-learning framework for unsupervised text generation. We define a heuristic scoring function that (roughly) estimates the quality of a candidate sentence for a task, and then perform stochastic local search (such as simulated annealing) to generate an output sentence. We also learn a sequence-to-sequence model that learns from the search results to improve inference efficiency and to smooth out search noise. Our search-and-learning framework shows high unsupervised performance in various natural language generation applications. Our technique should be useful in various industrial applications, especially for startups and the cold-start of new tasks.
What You’ll Learn:
The experience we had over the development
Dr. Lili Mou is an Assistant Professor at the Department of Computing Science, University of Alberta. He is also an Alberta Machine Intelligence Institute (Amii) Fellow and a Canada CIFAR AI (CCAI) Chair. Lili received his BS and PhD degrees in 2012 and 2017, respectively, from School of EECS, Peking University. After that, he worked as a postdoctoral fellow at the University of Waterloo and a research scientist at Adeptmind (a startup in Toronto, Canada). His research interests include deep learning applied to natural language processing as well as programming language processing. He has publications at top conferences and journals, including AAAI, ACL, CIKM, COLING, EMNLP, ICASSP, ICLR, ICML, IJCAI, INTERSPEECH, NAACL-HLT, NeruIPS, and TACL (in alphabetic order). He also has tutorials presented at EMNLP-IJCNLP’19 and ACL’20.
(Wednesday) 12:10 PM - 12:40 PM