Computer Science

News

Topic filter: All · Events · Awards · Jobs · Articles

Talk at MtHolyoke on Speech and Language Processing

An Introduction to TTI-Chicago, and Research in Speech and Language Processing

Karen Livescu  – Assistant Professor at TTI-Chicago
December 1st
4:00 p.m. – Kendade 305 

Abstract:

This talk will have two parts.  In the first, I will give an overview of TTI-Chicago (TTIC) and our PhD program.  In the second part, I will give a taste of current research at TTIC in the area of speech and language processing.  I will focus mainly on the topic of word embeddings, a way of representing a word using a vector of continuous-valued numbers.  We can think of each of the vector dimensions as representing some aspect of the word, and we can try to assign vectors to words so that similar words have similar vectors.  Word embeddings have been useful in a variety of natural language processing applications.  I will talk about both semantic word embeddings (where the vector is intended to represent the meaning of the word) and acoustic word embeddings (where the vector represents the way the word sounds).  I will address a number of questions related to this line of work:  How do we automatically learn word embeddings from data, such that they are useful for applications that we care about?  What do these vectors look like?  And how can we combine word embeddings to form, say, phrase or sentence vectors?

For a sneak peek, take a look at some relevant web pages:  

http://www.ttic.edu/phd-program.php

http://ttic.edu/livescu/SLATTIC/

Bio:

Karen Livescu is an Assistant Professor at TTI-Chicago.  She completed her PhD and post-doc in electrical engineering and computer science at MIT and her AB in Physics at Princeton University.  Karen’s main research interests are in speech and language processing, with a slant toward combining machine learning with linguistic structure.  Her recent work has included multi-view learning of speech and text representations, segmental neural models, and automatic sign language recognition.  She is a member of the IEEE Spoken Language Technical Committee, an associate editor for IEEE Transactions on Audio, Speech, and Language Processing, and a technical co-chair of ASRU 2015 and 2017.  She has organized/co-organized a number of recent workshops, including the ISCA SIGML workshops on Machine Learning in Speech and Language Processing and the Midwest Speech and Language Days.

November 29th, 2016 · Tags: 5 College, Talk