Seminars

Feb
10
Mon
Rada Mihalcea (University of Michigan) “Words and People” @ Hackerman Hall B17
Feb 10 @ 12:00 pm – 1:15 pm

Abstract

Language is not only about the words, it is also about the people. While much of the work in computational linguistics has focused almost exclusively on words (and their relations), recent research in the emerging field of computational sociolinguistics has shown that we can effectively leverage the close interplay between language and people. In this talk, I will explore this interaction, and show (1) that we can develop cross-cultural language models to identify words that are used in significantly different ways by speakers from different cultures; and (2) that we can effectively use information about the people behind the words to build better language representations.

Biography

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 computational linguistics, with a focus on lexical semantics, multilingual natural language processing, 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 ACL Vice-President. She is the recipient of a Presidential Early Career Award for Scientists and Engineers awarded by President Obama (2009) and she is an ACM Fellow (2019). In 2013, she was made an honorary citizen of her hometown of Cluj-Napoca, Romania.

Feb
17
Mon
Vivian Chen (National Taiwan University) “Scalable and Robust Conversational System” @ Hackerman Hall B17
Feb 17 @ 12:00 pm – 1:15 pm

Abstract

Even conversational systems have attracted a lot of attention recently, the current systems sometimes fail due to the errors from different components. This talk presents our recent research: 1) we focus on analyzing the bias in the SOTA models and learning language embeddings specifically for practical scenarios such as more noisy inputs during inference, and 2) secondly we investigate how to build scalable systems by leveraging the property of zero-shot learning or dual learning. Both directions enhance the robustness and scalability of conversational systems, showing the potential of guiding future research areas.

Biography

Yun-Nung (Vivian) Chen is currently an assistant professor in the Department of Computer Science & Information Engineering at National Taiwan University. She earned her Ph.D. degree from Carnegie Mellon University, where her research interests focus on spoken dialogue systems, language understanding, natural language processing, and multimodality. She received Google Faculty Research Awards, MOST Young Scholar Fellowship, FAOS Young Scholar Innovation Award, Student Best Paper Awards, and the Distinguished Master Thesis Award. Prior to joining National Taiwan University, she worked in the Deep Learning Technology Center at Microsoft Research Redmond.

Feb
21
Fri
Bhuvana Ramabhadran (Google) “Transliteration Based Approaches for Multilingual, Code-Switched Languages” @ Hackerman Hall B17
Feb 21 @ 12:00 pm – 1:15 pm

Abstract

Code-switching is a commonly occurring phenomenon in many multilingual communities, wherein a speaker switches between languages within a single utterance. Conventional Word Error Rate (WER) is not sufficient for measuring the performance of code-mixed languages due to ambiguities in transcription, misspellings and borrowing of words from two different writing systems. These rendering errors artificially inflate the WER of an Automated Speech Recognition (ASR) system and complicate its evaluation. Furthermore, these errors make it harder to accurately evaluate modeling errors originating from code-switched language and acoustic models.  Multilingual Automated Speech Recognition systems allow for the joint training of data-rich and data-scarce languages in a single model. This enables data and parameter sharing across languages, which is especially beneficial for the data-scarce languages. Most state-of-the-art multilingual models require the encoding of language information and therefore are not as flexible or scalable when expanding to newer languages.  Language independent multilingual models help to address this issue, and are also better suited for multicultural societies where several languages are frequently used together (but often rendered with different writing systems).

In this talk, I will discuss the use of a new metric, transliteration-optimized Word Error Rate (toWER) to evaluate ASR systems in code-switched languages. This metric smoothes out many of the irregularities by mapping all text to one writing system. I will also discuss  a new approach to building a language-agnostic multilingual ASR system which transforms all languages to one writing system through a many-to-one transliteration transducer. Thus, similar sounding acoustics are mapped to a single, canonical target sequence of graphemes, effectively separating the modeling and rendering problems. We show with Indic languages, that the language-agnostic multilingual model achieves up to 10% relative reduction in Word Error Rate (WER) over a language-dependent multilingual model.

Bio

Bhuvana Ramabhadran (IEEE Fellow, 2017, ISCA Fellow 2017) currently leads a team of researchers at Google, focusing on multilingual speech recognition and speech synthesis. Previously, she was a Distinguished Research Staff Member and Manager in IBM Research AI, at the IBM T. J. Watson Research Center, where she led a team of researchers in the Speech Technologies Group and coordinated activities across IBM’s world wide laboratories in the areas of speech recognition, synthesis, and spoken term detection. She has served as an elected member of the IEEE SPS Speech and Language Technical Committee (SLTC), and as Vice Chair and Chair (2014–2016),  served on the IEEE SPS conference board (2017-2018) and the editorial board of the IEEE Transactions on Audio, Speech, and Language Processing (2011–2015). She currently serves on the IEEE Flanagan Award Committee and is the Regional Director-At-Large for Region 6. She also serves on the ISCA board. Her research interests include speech recognition and synthesis algorithms, statistical modeling, signal processing, and machine learning. Some of her recent work has focused on understanding neural networks and methods to merge speech synthesis and recognition systems.

Feb
24
Mon
Tom Lippincott (JHU) “Computational Intelligence for the Humanities” @ Hackerman Hall B17
Feb 24 @ 12:00 pm – 1:15 pm

Abstract

A recurring task at the intersection of humanities and computational research is pairing data collected by a traditional scholar with an appropriate machine learning technique, ideally in a form that creates minimal burden on the scholar while yielding relevant, interpretable insights.

In this talk, I describe initial efforts to design a graph-aware autoencoding model of relational data that can be directly applied to a broad range of humanities research, and easily extended with improved neural (sub)architectures.  I then present results from an ongoing historical study of the post-Atlantic slave trade in Baltimore, illustrating several ways it can benefit traditional scholars. Finally, I briefly introduce a few additional historical and literary-critical studies, currently under-way in the Krieger school, that I hope to consider under the same framework in the coming year.

 

Feb
15
Mon
Eunsol Choi (University of Texas at Austin) “Learning to Understand Language in Context” @ via Zoom
Feb 15 @ 12:00 pm – 1:15 pm

Abstract

Many applications of natural language processing need to understand text from the rich context in which it occurs and present information in a new context. Interpreting the rich context of a sentence, either conversation history, social context, or preceding contents in the document, is challenging yet crucial to understand the sentence. In the first part of the talk, we study the context-reduction process by defining the problem of sentence decontextualization: taking a sentence together with its context and rewriting it to be interpretable out of context, while preserving its meaning. Typically a sentence taken out of a context is unintelligible, but decontextualization recovers key pieces of information and make sentences stand alone. We demonstrate the utility of this process, as a preprocessing for open-domain question answering and for generating an informative and concise answer to an information-seeking query. In the latter half of the talk, we focus on building models to integrate rich context to interpret single utterances more accurately. We study the challenges of interpreting rich context in question answering, by first integrating conversational history and by integrating entity information. Together, these works show how modeling interaction between text and the rich context in which it occurs can improve performances of NLP systems.

Biography

Eunsol Choi is an assistant professor in the computer science department at the University of Texas at Austin. Her research focuses on natural language processing, various ways to recover semantics from unstructured text. Recently, her research focused on question answering and entity analysis. Prior to UT, she was a visiting faculty researcher at Google AI. She received a Ph.D. from the University of Washington, working with Luke Zettlemoyer and Yejin Choi. She received an undergraduate degree in mathematics and computer science at Cornell University. She is a recipient Facebook Research Fellowship and has co-organized many workshops related to question answering at NLP and ML venues.

Feb
22
Mon
David Bamman (University of California, Berkeley) “Modeling the Spread of Information within Novels” @ via Zoom
Feb 22 @ 12:00 pm – 1:15 pm

Abstract

Understanding the ways in which information flows through social networks is important for questions of influence–including tracking the spread of cultural trends and disinformation and measuring shifts in public opinion.  Much work in this space has focused on networks where nodes, edges and information are all directly observed (such as Twitter accounts with explicit friend/follower edges and retweets as instances of propagation); in this talk, I will focus on the comparatively overlooked case of information propagation in *implicit* networks–where we seek to discover single instances of a message passing from person A to person B to person C, only given a depiction of their activity in text.

Literature in many ways presents an ideal domain for modeling information propagation described in text, since it depicts a largely closed universe in which characters interact and speak to each other.  At the same time, it poses several wholly distinct challenges–in particular, both the length of literary texts and the subtleties involved in extracting information from fictional works pose difficulties for NLP systems optimized for other domains.  In this talk, I will describe our work in measuring information propagation in these implicit networks, and detail an NLP pipeline for discovering it, focusing in detail on new datasets we have created for tagging characters and their coreference in text.  This is joint work with Matt Sims, Olivia Lewke, Anya Mansoor, Sejal Popat and Sheng Shen.

Biography

David Bamman is an assistant professor in the School of Information at UC Berkeley, where he works in the areas of natural language processing and cultural analytics, applying NLP and machine learning to empirical questions in the humanities and social sciences. His research focuses on improving the performance of NLP for underserved domains like literature (including LitBank and BookNLP) and exploring the affordances of empirical methods for the study of literature and culture. Before Berkeley, he received his PhD in the School of Computer Science at Carnegie Mellon University and was a senior researcher at the Perseus Project of Tufts University. Bamman’s work is supported by the National Endowment for the Humanities, National Science Foundation, an Amazon Research Award, and an NSF CAREER award.

Feb
26
Fri
Tuo Zhao (Georgia Tech) “Towards Automatic Evaluation of Dialog Systems: A Model-Free Off-Policy Evaluation Approach” @ via Zoom
Feb 26 @ 12:00 pm – 1:15 pm

Abstract

Reliable automatic evaluation of dialogue systems under an interactive environment has long been overdue. An ideal environment for evaluating dialog systems, also known as the Turing test, needs to involve human interaction, which is usually not affordable for large scale experiments. Though researchers have attempted to use metrics (e.g., perplexity, BLEU) in language generation tasks or some model-based reinforcement learning methods (e.g., self-play evaluation) for automatic evaluation, these methods only show very weak correlation with the actual human evaluation in practice.

To bridge such a gap, we propose a new framework named ENIGMA for estimating human evaluation scores based on recent advances of off-policy evaluation in reinforcement learning. ENIGMA only requires a handful of pre-collected experience data, and therefore does not involve human interaction with the target policy during the evaluation, making automatic evaluations feasible. More importantly, ENIGM is model-free and agnostic to the behavior policies for collecting the experience data, which significantly alleviates the technical difficulties of modeling complex dialogue environments and human behaviors. Our experiments show that ENIGMA significantly outperforms existing methods in terms of correlation with human evaluation scores.

Biography

Tuo Zhao (https://www2.isye.gatech.edu/~tzhao80/) is an assistant professor at Georgia Tech. He received his Ph.D. degree in Computer Science at Johns Hopkins University. His research mainly focuses on developing methodologies, algorithms and theories for machine learning, especially deep learning. He is also actively working in neural language models and open-source machine learning software for scientific data analysis. He has received several awards, including the winner of INDI ADHD-200 global competition, ASA best student paper award on statistical computing, INFORMS best paper award on data mining and Google faculty research award.

Center for Language and Speech Processing