Conferences, Lectures, & Seminars
Events for August
-
NL SEMINAR Improving Dialogue Agents with a Social Dimension and Understanding Monolingual Pre-Training for Bilingual Models
Thu, Aug 13, 2020 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Naitian Zhou and Omar Shaikh , ISI interns
Talk Title: Improving Dialogue Agents with a Social Dimension and Understanding Monolingual Pre-Training for Bilingual Models
Abstract: Naitian Zhou
TITLE Improving Dialogue Agents with a Social Dimension
The dialogue problem is challenging because a proper response must be conditioned on many different factors: knowledge about the language, knowledge about the world, knowledge about self, and knowledge about the speaker, to name a few. Prior research has focused on language modeling and "persona" modeling, encoding facts about the dialogue agent. In this project, I try to focus on an alternative dimension to dialogue: how do we condition the way we converse, based on our understanding of ourselves and our social relationship with our dialogue partner?
Omar Shaikh
TITLE Understanding Monolingual Pre Training for Bilingual Models
Monolingual embeddings from models like BERT are known to help on a variety of downstream tasks in a straightforward way. Usually, these embeddings are plug and play initializing models with BERT embeddings or using them as input representations result in increased model performance. However, supervised NMT tasks do not appear to benefit equally from traditional pretraining methods. We explore what makes NMT bilingual and BERT LM monolingual representations different on several probing tasks, and why certain training methods succeed in extracting performance from BERT embeddings from NMT tasks.
Biography: 1. Naitian Zhou is a rising junior at the University of Michigan studying computer science and data science. His interests include computational social science and natural language processing.
2. Omar is a Summer 2020 intern with Dr. Jon May. He is also a rising junior at Georgia Institute of Technology.
Host: Jon May and Emily Sheng
More Info: https://nlg.isi.edu/nl-seminar/
Webcast: https://youtu.be/QS3NVSbunt8Location: Information Science Institute (ISI) - Virtual Only
WebCast Link: https://youtu.be/QS3NVSbunt8
Audiences: Everyone Is Invited
Contact: Petet Zamar
Event Link: https://nlg.isi.edu/nl-seminar/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
NL Seminar-Qualitative Analysis of Unsupervised Neural Machine Translation
Thu, Aug 20, 2020 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Weiqiu You (ISI Intern), Univ. Of Pennsylvania
Talk Title: Qualitative Analysis of Unsupervised Neural Machine Translation
Series: NL Seminar
Abstract: Supervised neural machine translation SNMT models built on all the available parallel data result in higher BLEU on test sets than unsupervised neural machine translation UNMT models leveraging all available monolingual data.
Recently UNMT models such as XLM and MASS are reducing the gap with SNMT in terms of BLEU. Prior work has shown that linguistic and domain dissimilarity often hinder UNMTs performance. The question we're asking in this investigation is, when SNMT and UNMT do have comparable BLEU, do they exhibit qualitative differences, and if these differences can be detected by other metrics.
Biography: Weiqiu You is a rising first year PhD student at University of Pennsylvania who just graduated with MS from UMass Amherst. Her interests include machine translation and natural language processing in general.
Host: Jon May and Emily Sheng
More Info: https://nlg.isi.edu/nl-seminar/
Webcast: https://youtu.be/TawjeAeLbuALocation: Information Science Institute (ISI) - Virtual Only
WebCast Link: https://youtu.be/TawjeAeLbuA
Audiences: Everyone Is Invited
Contact: Petet Zamar
Event Link: https://nlg.isi.edu/nl-seminar/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor. -
NL Seminar- Translation of Asylum Testimonials from Low-Resource Languages
Thu, Aug 27, 2020 @ 11:00 AM - 12:00 PM
Information Sciences Institute
Conferences, Lectures, & Seminars
Speaker: Ugur Yavuz ISI Summer Intern, Dartmouth College
Talk Title: Translation of Asylum Testimonials from Low-Resource Languages
Abstract: Many asylum seekers along the southern border of the United States speak low-resource languages that are not available on commercial translation services. As a result, the translation of their testimonials poses a real challenge to the legal system, as well as non-governmental organizations. We will discuss potential techniques that would facilitate this task, such as transfer learning, domain adaptation, and corpus expansion, and also explain our work in compiling bilingual corpora for Mixtec and Kanjobal languages.
Biography: Ugur Yavuz is a summer intern working at the Natural Language Group with Dr. Jon May. He is a rising senior at Dartmouth College studying computer science and mathematics.
Host: Jon May and Emily Sheng
More Info: https://nlg.isi.edu/nl-seminar/
Webcast: https://youtu.be/xL9kMb2lTa4Location: Information Science Institute (ISI) - Virtual Only
WebCast Link: https://youtu.be/xL9kMb2lTa4
Audiences: Everyone Is Invited
Contact: Petet Zamar
Event Link: https://nlg.isi.edu/nl-seminar/
This event is open to all eligible individuals. USC Viterbi operates all of its activities consistent with the University's Notice of Non-Discrimination. Eligibility is not determined based on race, sex, ethnicity, sexual orientation, or any other prohibited factor.
