COMS 4705.H01: Natural Language Processing, Spring 2017
Time: Friday ? - ? pm
Place: ?
Professor:
Karl Stratos (Jang Sun Lee)
Teaching Assistants:
Christopher Hidey
Each week, TA office hours will be held at CLIC on Thursday 4-6pm.
Text Book: Speech and Language Processing, Second Edition, by Jurafsky and Martin. The second edition is currently on sale but we will be mostly using drafts of the third edition as it gets released by the authors.
Piazza: COMS W4705
Description | Prerequisites | Announcements | Assignments | Grading
Submission Policy | Collaboration Policy | Academic_Integrity
Grading Appeals | Email | Programming Languages | Important Dates | Students with Disabilities | Resources
This is a flipped version of COMS W4705. We will be using recorded video lectures by Dragomir Radev from previous year, and the instructor will hold regular help sessions to answer questions. Thus we will be covering similar topics, but assignments and exams are subject to change and new materials may be added.
Course description: Computational approaches to natural language processing and understanding. Recommended preparation: some previous or concurrent exposure to AI or Machine Learning, basic maturity in mathematics. The course has five major parts: 1. linguistic, mathematical, and computational background, 2. computational models of morphology, syntax, semantics, discourse, pragmatics, 3. core NLP technology: parsing, part of speech tagging, text generation, semantic analysis, etc., 4. applications: text classification, sentiment analysis, text summarization, question answering, machine translation, information extraction, etc., 5. neural networks. Programming exercises (in python) in several of these areas.
| Jan 17 | First Day of Class |
| March 13-17 | Spring Break |
| May 1 | Last Day of Class |
COMS W3134, W3136, or W3137; or the instructor's permission.
(20170110) Page created
(20170110) The SLP textbook and a CLIC account are required.
All assignments will be posted to Courseworks.
4 Homework Assignments 60%
Midterm 15%
Final 20%
Participation 5%
In the absence of a prior emailed authorization from the instructor,
you should turn in your assignments electronically by 11:59 PM on the
due date. For each day (or fraction of a day) that your submission is
late, it will be penalized 10%, for a maximum of 30%. After three
days, the assignment will be given a score of zero.
You will need to hand in the source code for the project, relevant
documentation, and a script of a test run of your program to show
that it actually works on the Clic machines.
Basic CLIC tutorial
The default grading criteria are:
Correctness 75%
Design/Structure 15%
Documentation 10%
We will usually provide a test suite with each assignment. You will
need to run your program on the test suite, and hand in a log of the
output. The test script will help us determine the correctness of your
program.
The documentation will help us assess the structure, design, and
efficiency of your program. Documentation need not be encyclopedic,
but should succinctly describe the structure and design of your
program. Normally, a page or so will suffice. Of course, we expect to
see comments in the source code as well.
You may discuss the course material and the textbook with other
students. You may also discuss the *requirements* of the
assignments. However, you cannot get help with the assignments and
exams themselves in oral or written form from anyone. If you are
unsure about this policy, ask the instructors.
Copying or paraphrasing someone's work (code included), or permitting your own work to be copied or paraphrased, even if only in part, is not allowed, and will result in an automatic grade of 0 for the entire assignment or exam in which the copying or paraphrasing was done. Your grade should reflect your own work. If you believe you are going to have trouble completing an assignment, please talk to the instructor or TAs in advance of the due date. For more information, please read this page very carefully.
If you have a question about your grade on a particular assignment (or
exam), write a short email to the TA in charge of that assignment.
Please submit any such requests within a week of receiving your grade.
Send your questions via email to the instructors mailing list. This
way it will reach the instructor and all TAs and you will get a reply
sooner.
The programming assignments will be in Python. You are expected to either know
Python already or to learn it on your own. Tutorial materials on
Python will be provided.
If you have any disability as defined under the Americans with
Disabilities Act and you believe that it may prevent you from
participating fully in class or it may make it difficult for you to
submit your assignments on time, please contact the instructor at the
beginning of the semester.
All resources about the class, including software, data sets, etc. are here.