Kelly at UW Computational Linguistics at University of Washington

6Jul/11Off

ACL was great, internship at Center for American Indian Languages

I survived my first ACL where I had the opportunity to present and get some great feedback on our work with formality in the Enron email corpus.  It was certainly an intimidating experience but I gained a lot from the comments of others in th field of natural language processing.

Now I am back in full swing at my internship at the University of Utah at the Center for American Indian Languages.  I'm currently working on some exciting projects for language documentation and revitalization.  I also have the opportunity to work on my Spanish as I've been building a few applications for Computer Aided Language Learning (CALL) en español.

I will be wrapping up this internship by the end of August so I have also begun my job search.  I will make a resume available here soon but if anyone knows of an opportunity for a computational linguist either based in Salt Lake City or remotely, I would love to hear about it.

14Jun/11Off

Enron email formality paper for ACL is now live

The Enron email formality paper is now live.  I look forward to presenting with Matt and Fei next week at ACL, but you can read the paper here now :

http://aclweb.org/anthology-new/W/W11/W11-0711.pdf

14Jun/11Off

Finished my coursework for CLMA

I can't believe how quickly the last year has passed.  I just finished my last quarter of coursework at UW.  It's true that I have learned a great deal, but I feel like I want to keep going.  There's simply too much to know in the field of Computational Linguistics.

This last quarter I took 3 classes:

One was an Independent Research project with Dr. Scott Farrar.  I hope to have more to share on this project very soon.  It's a very exciting Android tablet application for field linguists to store expressions in the field and also to help guide sessions with consultants.  Again, watch this space for more.

I also took a LING 575 seminar from Dr. Gina-Anne Levow on the topic of Discourse and Dialog.  There were many interesting papers and projects from the students in the seminar.  I worked on a paper with Nathan Imse proposing architectural and feature enhancements in recognizing troubles in the communication, particularly in recognition of anger and frustration.

Finally, the capstone course for the Master's program was LING 573.  This year as a project we worked as teams to build question-answering systems.  This was a great application of all the knowledge we have gained in the program and it was a lot of fun.  There was so much that could be done that it was difficult to choose how much to attack for a short 10 week course.  I had a great time working with Elliot Holt and I am very proud of the system that we delivered in the end.

As for now, I am doing an internship at the Center for American Indian Languages at the University of Utah where I got my undergrad in Computer Science.  So far I am loving the projects and the people I am working with.  I look forward to sharing more about what I learn and work on.

10May/11Off

Paper accepted to ACL 2011

The first paper I have submitted to a conference was accepted to ACL 2011 (June 19-24 in Portland, Oregon) to be part of a workshop on Language and Social Media.  You can see the list of accepted papers and topics here :

http://research.microsoft.com/en-us/events/lsm2011/default.aspx

I had the pleasure of working on this project with Dr. Fei Xia and Matt Hohensee.

This work is a case study in the Enron email corpus which aims to show how techniques of natural language processing can be used to study theories of sociolinguistics.  It has several interesting applications which are possible so  I hope that this scratch in the surface will be able to continue  with more research.

28Jan/11Off

Update on Winter Quarter 2011

So far I am 4 weeks into Winter quarter and it is very intense.  Since my background is in Computer Science I am excited to be doing so much more programming but it's also very time intensive.   I'm currently taking the following 3 classes :

LING 571 - Deep Processing Techniques for Natural Language Processing

LING 572 - Statistical Methods in Natural Language Processing

LING 575 - Computational Linguistics and Social Media

So far this quarter is much more demanding than the Fall but it's also very rewarding.  In particular, I am very excited about the research we are doing with Dr. Fei Xia in LING 575.  We are using an email dataset from Enron and dealing with some very interesting topics.  I am thrilled about my topic and can't wait to find out if some of the correlations I expect to find in email behavior will turn out to be correct.  I am looking forward to the results so much that I am tempted to get my software system done early so that I can start analyzing the results.

29Dec/10Off

Finished Autumn Quarter 2010

Waveform of Kirundi in Sound Forge. I spent lots of hours listening to this and transcribing the language into the International Phonetic Alphabet...

So I have officially wrapped up my first quarter of grad school.  It was not easy, but I am very happy about what I have learned and what I have accomplished so far.  This quarter was a mix of several different classes.  Each presented its own unique challenges that helped me to grow.

Ling 566 (Introduction to Syntax for Computational Linguistics) is probably the course that stretched me most.  It was exciting to perform a complex model of syntax that not only can be implemented on a computer but is also highly aligned with research in psycholinguistics.  The textbook was excellent and it was very rewarding to continually build up from a simple grammar into a complex one which is capable of modeling language variation and phenomenon such as Long Distance Dependencies.  This was very difficult for me, so I am glad I put extra time into this course.

Ling 570 (Shallow Processing Techniques for NLP) was less of a challenge for me personally but the material was rewarding as we were able to build systems to perform Part-of-speech tagging, morphotactic expansion, document classification and several others.  I look forward to the continuation of this series with 571 and 572 this Winter.

Ling 550 (Introduction to Linguistics Phonetics) was a treat.  Since my background is in Computer Science, it was a thrill to take an actual graduate Linguistics course.  I can't believe how much material we covered over a 10 week course and how much I know am aware how we generate and hear sounds to communicate with one another.  Of course since I love to learn languages it was great to learn more formally about how certain sounds are made since before I would simply imitate what I heard from others but now that I am more aware of the articulatory processes involved in making sounds, I am much more careful in speaking French and Spanish and I am much better equipped to learn other languages I would like to study like Portugese and Arabic.  I learned how about the different sounds of the world's languages including clicks and implosives and often found myself practicing them while walking around Salt Lake which made me look crazier than I am.  The real reward in this class was using all the skills to perform research and make a real recording of a language I didn't know just as a field linguist would.  I chose Kirundi since I have taught English to several refugees from Burundi.  My final paper and recording turned out much better that I could have hoped and now I intend to share my work.

I am considering sharing my paper with the Journal of the International Phonetic Association since much of what I found about the language I had to translate from French sources.

In addition, I plan to augment the current page for Kirundi on Wikipedia with the research I found and include the sources which were helpful to me in studying the language.  Check the page now and check it in a week and hopefully I will have added quite a bit to the page.

I have also shared the .PDF of the final paper and the audio files of the recording over at this new Kirundi page.

All in all, this was a great quarter and I am looking forward to starting up again next week and I will be in Seattle to start the first week of classes.

23Sep/10Off

Starting Autumn Quarter 2010

Tomorrow I am flying to Seattle to attend Orientation for the CLMA program and meet many people who I have so far maintained contact only via e-mail.

It's been years since I've been to Seattle although at one time I knew it fairly well since I would visit twice a year while I worked for Microsoft.

Besides orientation and meeting everyone, I will also be on campus for the first week of class. I have a busy quarter ahead since I am studying full-time to hopefully complete my coursework in one year.  Here are the classes I am taking this Autumn :

LING 450 - Introduction to Linguistics Phonetics

LING 566 - Introduction to Syntax for Computational Linguistics

LING 570 - Shallow Processing Techniques for NLP

23Sep/10Off

Just finished LING 473

I just recently wrapped up work on LING 473 - Computational Linguistics Fundamentals.  This was a great class and I highly recommend it to any student entering the CLMA program.  I was hesitant to register initially, but I was very pleasantly surprised with the curriculum and the work we were able to accomplish by the end of the quarter.

As I mentioned, I was initially hesitant since I had previously seen the course billed as a "refresher" course.  To me I read this as meaning that there would be a lot of presentation of material and less hands-on projects.  While we did cover a lot of material, our knowledge was solidified by the great projects we were assigned by our instructor Glenn Slayden.

I am also impressed with how well the distance learning is working out so far since I do not attend class in Seattle (I live in Salt Lake City).  The Adobe Connect software was great for real-time interaction and I found the fact that the classes were recorded to be particularly handy since I was able to review parts of the material as many times as I needed.

I feel very prepared to enter the CLMA program and I am looking forward to what I will learn next.

30Jul/10Off

Getting started in the program

I just got started in the program with LING 473 - Computational Linguistics Fundamentals.  I am really enjoying the course so far.  It's been a long time since I have been in school, but I am completely thrilled by new problems to solve and the exciting field of Computational Linguistics.

I am looking forward to visiting Seattle in September to meet the staff and students who I have so far been in contact with online.

12May/10Off

Accepted to UW

I was just accepted to the CLMA program at UW.

I couldn't be happier and I can't wait to get to work!