The following speakers have graciously accepted to give keynotes at NAACL 2019. The titles and abstracts will be announced soon.
Title: When the Computers Spot the Lie (and People Don’t)
Abstract: Whether we like it or not, deception occurs everyday and everywhere: thousands of trials take place daily around the world; little white lies: “I’m busy that day!” even if your calendar is blank; news “with a twist” (a.k.a. fake news) meant to attract the readers attention or influence people in their future undertakings; misinformation in health social media posts; portrayed identities, on dating sites and elsewhere. Can a computer automatically detect deception in written accounts or in video recordings? In this talk, I will overview a decade of research in building linguistic and multimodal resources and algorithms for deception detection, targeting deceptive statements, trial videos, fake news, identity deception, and health misinformation. I will also show how these algorithms can provide insights into what makes a good lie - and thus teach us how we can spot a liar. As it turns out, computers can be trained to identify lies in many different contexts, and they can often do it better than humans do.
Rada Mihalcea is a Professor of Computer Science and Engineering at the University of Michigan and the Director of the Michigan Artificial Intelligence Lab. Her research interests are in lexical semantics, multilingual NLP, and computational social sciences. She serves or has served on the editorial boards of the Journals of Computational Linguistics, Language Resources and Evaluations, Natural Language Engineering, Journal of Artificial Intelligence Research, IEEE Transactions on Affective Computing, and Transactions of the Association for Computational Linguistics. She was a program co-chair for EMNLP 2009 and ACL 2011, and a general chair for NAACL 2015 and *SEM 2019. She currently serves as the ACL Vice-President Elect. She is the recipient of an NSF CAREER award (2008) and a Presidential Early Career Award for Scientists and Engineers awarded by President Obama (2009). In 2013, she was made an honorary citizen of her hometown of Cluj-Napoca, Romania.
Title: Leaving the Lab: Building NLP Applications that Real People can Use
Abstract: There is a chasm between an NLP technology that works well in the research lab and something that works for applications that real people use. Research conditions are often theoretical or idealized. The first time they contribute to industry projects, many theoretical researchers are surprised to discover how much goes into building outside the lab, and how hard it is to build data products for real people ethically and transparently. This talk explores my NLP journey in three stages: working as an academic NLP researcher, learning to be a practical creator of NLP products in industry, and becoming the founding CEO of an NLP business. While each role has used my background in computational linguistics in essential ways, every step has also required me to learn and unlearn new things along the way. The further I have gone in my industry career, the more critical it has become to define and work within a well-established set of principles for data ethics. This talk is for academic researchers considering industry careers or collaborations, for people in industry who started out in academia, and for anyone on either side of the divide who wants to make NLP products that real people can use
Kieran Snyder is the CEO and Co-Founder of Textio, the augmented writing platform. For anything you write, Textio tells you ahead of time who’s going to respond based on the language you’ve used. Textio’s augmented writing engine is designed to attach to any large text corpus with outcomes to find the patterns that work. Prior to founding Textio, Kieran held product leadership roles at Microsoft and Amazon. Kieran has a PhD in linguistics from the University of Pennsylvania. Her work has appeared in Fortune, Re/code, Slate, and the Washington Post.
Title: Data as a Mirror of Society: Lessons from the Emerging Science of Fairness in Machine Learning
Abstract: Language corpora reflect human society, including cultural stereotypes, prejudices, and historical patterns. By default, statistical language models will absorb these stereotypes. As a result, NLP systems for word analogy generation, toxicity detection, and many other tasks have been found to reflect racial and gender biases. Based on this observation, I will discuss two emerging research directions. First, a deeper understanding of human culture can help identify possible harmful stereotypes in algorithmic systems. The second research direction is the converse of the first: if data is a mirror of society, machine learning can be used as a magnifying lens to study human culture.
Arvind Narayanan is an Associate Professor of Computer Science at Princeton. His research has shown how state-of-the-art word embeddings reflect racial, gender, and other cultural stereotypes. He leads the Princeton Web Transparency and Accountability Project to uncover how companies collect and use our personal information. His doctoral research showed the fundamental limits of de-identification, for which he received the Privacy Enhancing Technologies Award. Narayanan also co-created a Massive Open Online Course as well as a textbook on Bitcoin and cryptocurrency technologies.