Advanced Natural Language Processing

CS 690N, Spring 2016, UMass Amherst

Instructor: Brendan O'Connor

Override waitlist

The course is currently full. If you are still interested please sign up on the override waitlist here: www.cs.umass.edu/overrides. Make sure to carefully describe your reason to take the course. We will read it. Demand is high and we will not be able to admit everyone who wants to enroll. PhD students will be given priority, since CS 585 already exists for non-PhD students.

Course Information

Description: This course covers a broad range of advanced level topics in natural language processing. It is intended for graduate students in computer science who have familiarity with machine learning fundamentals. It may also be appropriate for computationally sophisticated students in linguistics and related areas. Topics include probabilistic models of language, computationally tractable linguistic representations for syntax and semantics, neural network models for language, and selected topics in discourse and text mining. After completing the course, students should be able to read and evaluate current NLP research papers. Coursework includes homework assignments and a final project.

Prerequisites: Familiarity with multivariate calculus, probability theory, dynamic programming, and implementation of machine learning algorithms, such as from COMPSCI 585, 688, 689, STAT 697ML, or equivalent. Previous experience in linguistics or natural language processing may be helpful, but is not required.

Students: The course is open to Computer Science and Linguistics PhD students. Other students may be admitted depending on circumstances. We expect we will not be able to admit all interested students, sorry!

Should I take both 585 and 690N? We are recommending against it for most students. First, there is some overlap in topics, so time may be wasted if you already took 585. Second, there is considerably more advanced machine learning, which 585 does not entirely prepare students for (though it helps a little bit). Third, space in the course is reserved for PhD students (since 585 already exists), so we do not know if there will be space to accommodate all interested students. All that said, there may be some cases where taking both 585 and 690N makes sense; please ask us if you have questions.

Topics: see draft syllabus.

Readings: Readings will include research papers and may include selections from:

Related courses

See this list of courses in NLP and related areas offered at UMass and nearby.

Courses elsewehre

Georgetown, Stanford, Johns Hopkins, Georgia Tech, U Texas, Berkeley, Coursera now discontinued?, etc.