|Emily M. Bender||Gita Dhungana|
|Office Hours:||Tuesdays 3-4, Thursdays 1-2||Mondays 10-11, Wednesdays 10-11|
|Office Location:||GUG 414-C; Zoom (link in Canvas)||Zoom (link in Canvas)|
|Email:||ebender at uw||gitad at uw|
Learning outcomes. By the end of this course, students will:
Computational linguistics is a broad field incorporating research and techniques for processing language with computers at all levels of linguistic structure. In this class, we will survey various topics and tasks in computational linguistics focusing on linguistic structure. While we will cover some of the basics of Natural Language Processing (which we will consider a separate subfield), this class will not focus on one specific approach (such as deep learning). Students in this class are expected to have a background in either computer science or linguistics, but not necessarily both. Expect this class to be difficult at times and easy at others. We hope to offer something new and interesting for everyone.
We welcome and embrace students of all ethnic and religious backgrounds and with all abilities in this class. We want this quarter to be a welcoming and positive experience for everyone. As such, it is both our personal policy and the policy of the University of Washington to create inclusive and accessible learning environments. Below are the university’s statements regarding accessibility and accommodations. We also invite you to share with us any needs you might have that are undocumented or not covered by these policies.
Your experience in this class is important to me. It is the policy and practice of the University of Washington to create inclusive and accessible learning environments consistent with federal and state law. If you have already established accommodations with Disability Resources for Students (DRS), please activate your accommodations via myDRS so we can discuss how they will be implemented in this course.
If you have not yet established services through DRS, but have a temporary health condition or permanent disability that requires accommodations (conditions include but not limited to; mental health, attention-related, learning, vision, hearing, physical or health impacts), contact DRS directly to set up an Access Plan. DRS facilitates the interactive process that establishes reasonable accommodations. Contact DRS at disability.uw.edu
Washington state law requires that UW develop a policy for accommodation of student absences or significant hardship due to reasons of faith or conscience, or for organized religious activities. The UW's policy, including more information about how to request an accommodation, is available at Faculty Syllabus Guidelines and Resources. Accommodations must be requested within the first two weeks of this course using the Religious Accommodations Request form available at https://registrar.washington.edu/students/religious-accommodations-request/.
COVID 19 policy: Class will meet in person and also be accessible via Zoom. I strongly encourage you to come to class in person as much as possible, but if you are unwell, please attend via Zoom. I will be wearing a mask and invite you to do the same. I will also do my best to make sure the classroom is well ventilated. (For the week of May 8, I will be traveling and so class will be only on Zoom.)
Students are expected to complete the assigned readings before each lecture. Lecture and Lab/Section will connect with the readings, but not everything in the readings will be covered in lecture. Homework assignments and exams may nonetheless cover material in the readings not gone over in class.
All homework assignments and the final project will include a significant writing component, weight at or near 1/2 of the assignment grade. Be sure to save time to do a careful job on your write up.
We expect all write ups to be turned in as pdf files, even if they started as plain text files that we gave you.
Collaboration policy: Students are encouraged to work with each other on the homework, both in small groups and by posting & answering questions on Canvas. However, each student must turn in their own answers (both code and write up). No copying or sharing code or prose is allowed. Also, students who have collaborated must acknowledge the collaboration in their write ups (e.g. "I discussed this problem with [name] on Canvas as we were working on it." or "I pair programmed with [name].").
Plagiarism policy: Plagiarism is strictly forbidden. The offender will get 0 points for the plagiarized assignment and will be reported to the University. Note: It is very easy to detect not only plagiarized text but also a (piece of a) program, or even a mathematical solution that was adapted from something posted on the internet. Just don't. Submit your own solution, and rest assured, it will be unique! Note2: We also consider the output of text synthesis machines (like Chat GPT) to be plagiarism. Please do not waste our time with synthetic text.
Late homework policy: Unless prior arrangements are made, homework turned in late but within 24 hours of the deadline will be graded at 80% credit, homework turned between 24 and 48 hours will be graded at 70% credit, and homework turned in later than that will not be graded. No late final projects or reading questions will be accepted.
Grades will be based on (this may be updated through the start of class):
|3/30||Regular expressions||JM Ch 2 (through 2.4)|
|3/31||Slides||Intro survey, Assignment 0|
|4/4||Dialogue Systems and Chatbots||JM Ch 15.1-15.3, 15.6||15.4, 15.5, BL #5-7, BL Ch 9|
|4/6||Finite state methods in phonology and morphology||Bender 2013, Ch 3; FOMA tutorial||Karttunen 1998|
|4/11||Machine Learning, Bird's Eye View||Mitchel 2017||Pilehvar and Camacho-Collados (in prep), Ch 2.2|
|4/13||Evaluation and Error Analysis||Resnik & Lin 2010, Kummerfeld et al 2012||Fokkens et al 2013, Wu et al 2019,van Miltenburg et al 2021|
|4/18||Societal Impact of NLP||Nathan et al 2007||Bender, Gebru et al 2021,Blodgett et al 2020,Gonen and Goldberg 2019, Sap et al 2019|
|4/20||Data and Model Documentation||Bender and Friedman 2018||Bender et al 2021,Gebru et al 2021, Mitchell et al 2019|
|4/21||slides||Project Milestone 1|
|4/25||N-gram Language Modeling||JM Ch 3 (through 3.6)||JM 3.7|
|4/27||Neural LMs||JM Ch 7 (primarily 7.5)||Devlin et al 2019|
|5/2||PCFGs||JM Ch 17 (can skip 17.7), JM Appendix C, C1-C4||JM C5, JM C6|
|5/4||Dependency Parsing||JM Ch 18.1-18.3||JM 18.4, JM 18.5, Nivre et al 2016|
|5/9||Logical Representation of Sentence Meaning||JM Ch 19.1-19.3||JM 19.4, JM 19.5, BL Ch 7, BL Ch 8|
|5/11||Grammar-Based Treebanking,||Flickinger et al 2017 [audio]||Bender et al 2015, Buys and Blunsom 2017, Chen et al 2018|
|5/12||slides||Project Milestone 2|
|5/16||Vector Semantics||JM Ch 6 (through 6.6)||JM 6.7, BL #25-26|
|5/18||Word Embeddings J&M slides, local slides||JM Ch 6.8-6.13||Schluter 2018, Bloem et al 2019,Pilehvar and Camacho-Collados (2020), Ch 3|
|5/23||>Linguistic Semantics and NLP||Bender and Koller 2020 ([Audio version])||Dua et al 2019, Niven and Kao 2019|
|5/25||The Grammar Matrix and AGGREGATION|
|5/30||Catch-up/review/wrap-up + Black Mirror Writer's Room|
|6/1||Term Project Presentations|
|6/2||Term Project Presentations||Reflection 2|
|TUESDAY 6/6||Final projects due|