CSCI-GA.2590 -- Natural Language Processing -- Spring 2013
The final exam
- Worth 30 points towards final grade
- Given Tuesday, May 21st, 2013,
5:00 -- 6:50, 202 Warren Weaver Hall (usual class day, time, place)
- Closed book but
- You may bring one or two sheets of notes, double sided, with
your name on each sheet. (This may be convenient for
definitions or formulas, for example.)
- You may bring a simple calculator, which may be helpful for
questions on HMM or PCFG probabilities or word similarities.
No other electronic equipment is permitted.
- Approximately 6 to 8 questions
Most questions will be of one of the following types.
Many of these correspond directly to questions asked for homework.
In addition, there may be a few short answer questions corresponding
to major points of a lecture; some of these are included in the list
below. I may also ask a short (few sentence) essay question
about an issue we have discussed in the lectures.
- English sentence structure: Label the constituents (NP,
VP, PP, etc.) of an English sentence based on the grammar given in
Chapter #12 (and summarized in the handout for homework #1). If
the sentence is ambiguous, show its multiple parses. If the
sentence violates some grammatical constraint, describe the constraint.
(lecture #2, homework #1).
- Context-free grammar: Extend the context-free grammar to
cover an additional construct, or to capture a grammatical constraint.
- Parsing: Given a very small context-free grammar, to step
through the operation, or count the number of operations performed by a
top-down backtracking parser, a bottom-up parser, or a chart parser (homework
#2). What is the [time] complexity of these parsers? Convert the
constituent structure into a dependency structure.(lecture #3)
- POS tagging: Tag a sentence using the Penn POS tags
- HMMs and the Viterbi decoder: Describe how POS tagging can
be performed using a probabilistic model (J&M sec. 5.5 and chap 6;
Create an HMM from some POS-tagged training data. Trace the operation
of a Viterbi decoder. Compute the likelihood of a given tag sequence and
the likelihood of generating a given sentence from an HMM (homework #3).
What is the [time] complexity of the decoder?
- Chunkers and name taggers. Explain how BIO tags can
be used to reduce chunking or name identification to a token-tagging task.
Explain how chunking can be evaluated. (lecture #6).
- Maximum entropy: Explain how a maximum-entropy model
can be used for tagging or chunking (lecture #6 and homework #6).
Suggest some suitable features for each task.
- Jet: be able to extend, or trace the operation, of one of
the Jet pattern sets we have distributed and discussed (for noun and verb
groups, and for appointment events). Analyze and correct a shortcoming in
the appointment patterns (homework #8).
- Lexical semantics and word sense disambiguation:
given two words, state their semantic relationship;
given a word with two senses and small training set of contexts for
each of the two senses, apply the naive Bayes procedure to resolve the
sense of the word in a test case (J&M 20.2.2); given two words
and a few sentences containing them, compute their cosine
similarity (lecture #8).
- Reference resolution:
analyze a reference resolution problem -- identify the type of anaphora
and the constraints and preferences which would lead a system to select
the correct antecedent (lecture #10).
- Probabilistic CFG: Train a probabilistic CFG from some
parses; apply this PCFG to disambiguate a sentence. Explain how this PCFG can
be extended to capture lexical information. Compute
lexically-conditioned probabilities. (homework #8)
- Machine translation.Give the basic formula for noisy
channel translation. Explain how an n-gram language model can be computed.
What assumption is made by IBM Model 1?