N-grams and Smoothing

CSC 371: Spring 2012

 Last lecture
• Uncertainty
– How to represent uncertainty
• Probability and its axioms
– As degree of agent’s belief
• Joint distribution
– P(A,B,C,D … Z)
• Bayes rule
– P(X,Y) = P(X|Y) P(Y) = P(Y/X) P(X)
• Independence
– 2 random variables A and B are independent iff
P(a, b) = P(a) P(b) for all values a, b

• Conditional independence
– 2 random variables A and B are conditionally independent given C iff
P(a, b | c) = P(a | c) P(b | c) for all values a, b, c
 Today’s Lecture
• Use probability axioms to model word prediction
• N-grams
• Smoothing

• Reading:
– Sections 4.1, 4.2, 4.3, 4.5.1 in Speech and Language
Processing [Jurafsky & Martin]
 Next Word Prediction
• From a NY Times story...
– Stocks ...
– Stocks plunged this ….
– Stocks plunged this morning, despite a cut in interest
rates
– Stocks plunged this morning, despite a cut in interest
rates by the Federal Reserve, as Wall ...
– Stocks plunged this morning, despite a cut in interest
rates by the Federal Reserve, as Wall Street began
– Stocks plunged this morning, despite a cut in
interest rates by the Federal Reserve, as Wall
Street began trading for the first time since last …
– Stocks plunged this morning, despite a cut in
interest rates by the Federal Reserve, as Wall
Street began trading for the first time since last
Tuesday's terrorist attacks.
 Human Word Prediction
• Clearly, at least some of us have the ability to
predict future words in an utterance.
• How?
– Domain knowledge: red blood vs. red hat
– Syntactic knowledge: the…<adj|noun>
– Lexical knowledge: baked <potato vs. chicken>
 Claim
• A useful part of the knowledge needed to allow
Word Prediction can be captured using simple
statistical techniques
• In particular, we'll be interested in the notion of the
probability of a sequence (of letters, words,…)
 Useful Applications
• Why do we want to predict a word, given some
preceding words?
– Rank the likelihood of sequences containing various
alternative hypotheses, e.g. for ASR
Theatre owners say popcorn/unicorn sales have doubled...
– Assess the likelihood/goodness of a sentence, e.g. for
text generation or machine translation
 Applications of word prediction
• Spelling checkers
– They are leaving in about fifteen minuets to go to her house
– The study was conducted mainly be John Black.
– I need to notified the bank of this problem.

• Mobile phone texting

– He is trying to fine out.
– Hopefully, all with continue smoothly in my absence.

• Speech recognition
– Theatre owners say popcorn/unicorn sales have doubled...

• Handwriting recognition
• Disabled users
• Machine Translation
September 2003 9
 N-Gram Models of Language
• Use the previous N-1 words in a sequence to
predict the next word
• Language Model (LM)
– unigrams, bigrams, trigrams,…
• How do we train these models?
– Very large corpora
 Corpora
• Corpora are online collections of text and speech
– Brown Corpus
– Wall Street Journal
– AP newswire
– Hansards
– Timit
– DARPA/NIST text/speech corpora (Call Home, Call
Friend, ATIS, Switchboard, Broadcast News, Broadcast
Conversation, TDT, Communicator)
– TRAINS, Boston Radio News Corpus
 Terminology

• Sentence: unit of written language

• Utterance: unit of spoken language
• Word Form: the inflected form as it actually
appears in the corpus
• Lemma: an abstract form, shared by word forms
having the same stem, part of speech, word
sense – stands for the class of words with same
stem
– chairs [chair] , reflected [reflect], ate [eat],
• Types: number of distinct words in a corpus
(vocabulary size)
• Tokens: total number of words
 Simple N-Grams
• Assume a language has T word types in its lexicon,
how likely is word x to follow word y?
– Simplest model of word probability: 1/T
– Alternative 1: estimate likelihood of x occurring in new text
based on its general frequency of occurrence estimated
from a corpus (unigram probability)
popcorn is more likely to occur than unicorn
– Alternative 2: condition the likelihood of x occurring in the
context of previous words (bigrams, trigrams,…)
mythical unicorn is more likely than mythical popcorn
 Computing the Probability of a
Word Sequence
• By the Chain Rule we can decompose a joint probability, e.g.
P(w1,w2,w3) as follows
P(w1,w2, ...,wn) = P(wn|wn-1,wn-2,...,w1) P(wn-1|wn-2, ...,w1) … P(w2|w1) P(w1)
P(wn-1, ...,w1)
n
P(wn) P(wk | w1k 1)
1 k 1

• Compute the product of component conditional probabilities?

– P(the mythical unicorn) = P(the) * P(mythical|the) * P(unicorn|the
mythical)

• But…the longer the sequence, the less likely we are to find it

in a training corpus
P(Most biologists and folklore specialists believe that in fact the mythical
unicorn horns derived from the narwhal)

• What can we do?

 Bigram Model
• Approximate by
– E.g., P(unicorn|the mythical) by P(unicorn|mythical)
P(wn |w1n 1) P(wn | wn 1)
• Markov assumption: the probability of a word
depends only on the probability of a limited history
• Generalization: the probability of a word
depends only on the probability of the n
previous words
– trigrams, 4-grams, 5-grams,…
– the higher n is, the more data needed to train
– backoff models…
 Using N-Grams

• For N-gram models

– P(wn | w1n 1) P(wn |wnn 1N 1)
• E.g. for bigrams,
P(w8| w18 1) P(w8| w88 21 1)

– By the Chain Rule we can decompose a joint

probability, e.g. P(w1,w2,w3) as follows
P(w1,w2, ...,wn) = P(wn|wn-1,wn-2,...,w1) P(wn-1|wn-2, ...,w1) …
P(w2|w1) P(w1)
n
P(wn) P(wk | wkk 1 )
1 k 1
N 1
• For bigrams then, the probability of a sequence is
just the product of the conditional probabilities of its
bigrams, e.g.
P(the,mythical,unicorn) = P(unicorn|mythical) P(mythical|the)
P(the|<start>)
 Training and Testing
• N-Gram probabilities come from a training corpus
– overly narrow corpus: probabilities don't generalize
– overly general corpus: probabilities don't reflect task or
domain
• A separate test corpus is used to evaluate the
model
– held out test set; development (dev) test set
– Simple baseline
– results tested for statistical significance – how do they
differ from a baseline? Other results?
 A Simple Bigram Example
• Estimate the likelihood of the sentence I want to eat
Chinese food.
– P(I want to eat Chinese food) = P(I | <start>) P(want | I)
P(to | want) P(eat | to) P(Chinese | eat) P(food | Chinese)
P(<end>|food)
• What do we need to calculate these likelihoods?
– Bigram probabilities for each word pair sequence in the
sentence
– Calculated from a large corpus
 Early Bigram Probabilities
from BERP
Eat on .16 Eat Thai .03

Eat some .06 Eat breakfast .03

Eat lunch .06 Eat in .02

Eat dinner .05 Eat Chinese .02

Eat at .04 Eat Mexican .02

Eat a .04 Eat tomorrow .01

Eat Indian .04 Eat dessert .007

Eat today .03 Eat British .001

<start> I .25 Want some .04
<start> I’d .06 Want Thai .01
<start> Tell .04 To eat .26
<start> I’m .02 To have .14
I want .32 To spend .09
I would .29 To be .02
I don’t .08 British food .60
I have .04 British restaurant .15
Want to .65 British cuisine .01

Want a .05 British lunch .01

• P(I want to eat British food) = P(I|<start>)
P(want|I) P(to|want) P(eat|to) P(British|eat)
P(food|British) = .25*.32*.65*.26*.001*.60 =
.000080
– Suppose P(<end>|food) = .2?
– How would we calculate I want to eat Chinese
food ?
• Probabilities roughly capture ``syntactic'' facts
and ``world knowledge''
– eat is often followed by an NP
– British food is not too popular
• N-gram models can be trained by counting
and normalization
 Early BERP Bigram Counts
I Want To Eat Chinese Food lunch

I 8 1087 0 13 0 0 0

Want 3 0 786 0 6 8 6

To 3 0 10 860 3 0 12

Eat 0 0 2 0 19 2 52

Chinese 2 0 0 0 0 120 1

Food 19 0 17 0 0 0 0

Lunch 4 0 0 0 0 1 0
 Early BERP Bigram
Probabilities
• Normalization: divide each row's counts by
appropriate unigram counts for wn-1
I Want To Eat Chinese Food Lunch

3437 1215 3256 938 213 1506 459

• Computing the bigram probability of I I

– C(I,I)/C( I in call contexts )
– p (I|I) = 8 / 3437 = .0023
• Maximum Likelihood Estimation (MLE):
relative frequencyfreq (w1, w2)
freq(w1)
 What do we learn about the
language?
• What's being captured with ...
– P(want | I) = .32
– P(to | want) = .65
– P(eat | to) = .26
– P(food | Chinese) = .56
– P(lunch | eat) = .055
• What about...
– P(I | I) = .0023
– P(I | want) = .0025
– P(I | food) = .013
– P(I | I) = .0023 I I I I want
– P(I | want) = .0025 I want I want
– P(I | food) = .013 the kind of food I want is ...
 Approximating Shakespeare

• Generating sentences with random

unigrams...
– Every enter now severally so, let
– Hill he late speaks; or! a more to leg less first you
enter
• With bigrams...
– What means, sir. I confess she? then all sorts, he
is trim, captain.
– Why dost stand forth thy canopy, forsooth; he is
this palpable hit the King Henry.
• Trigrams
– Sweet prince, Falstaff shall die.
– This shall forbid it should be branded, if renown
made it empty.
• Quadrigrams
– What! I will go seek the traitor Gloucester.
– Will you not tell me who I am?
– What's coming out here looks like Shakespeare
because it is Shakespeare
• Note: As we increase the value of N, the
accuracy of an n-gram model increases,
since choice of next word becomes
increasingly constrained
 N-Gram Training Sensitivity
• If we repeated the Shakespeare experiment but
trained our n-grams on a Wall Street Journal
corpus, what would we get?
• Note: This question has major implications for
corpus selection or design
WSJ is not Shakespeare: Sentences Generated
from WSJ
 Some Useful Observations

• There are 884,647 tokens, with 29,066 word form

types, in an approximately one million word
Shakespeare corpus
– Shakespeare produced 300,000 bigram types out of 844
million possible bigrams: so, 99.96% of the possible
bigrams were never seen (have zero entries in the table)
• A small number of events occur with high frequency
• A large number of events occur with low frequency
• You can quickly collect statistics on the high
frequency events
• You might have to wait an arbitrarily long time to get
valid statistics on low frequency events
• Some zeroes in the table are really zeros But others
are simply low frequency events you haven't seen
yet. How to address?
 Smoothing
• Words follow a Zipfian distribution
– Small number of words occur very frequently
– A large number are seen only once
– Zipf’s law: a word’s frequency is approximately inversely
proportional to its rank in the word distribution list
• Zero probabilities on one bigram cause a zero
probability on the entire sentence
• So….how do we estimate the likelihood of unseen
n-grams?

32
Slide from Dan Klein
 Laplace Smoothing

• For unigrams:
–Add 1 to every word (type) count to get an
adjusted count c*
–Normalize by N (#tokens) + V (#types)
–Original unigram probability
P(w ) c i
i

N
–New unigram probability
P (w )
LP i
c 1
i

N V
 Unigram Smoothing
Example
P (w ) c 1
• Tiny Corpus, V=4; N=20
i
LP i

N V
Word True Ct Unigram New Ct Adjusted
Prob Prob
eat 10 .5 11 .46

British 4 .2 5 .21

food 6 .3 7 .29

happily 0 .0 1 .04

20 1.0 ~20 1.0

• So, we lower some (larger) observed
counts in order to include
unobserved vocabulary
• For bigrams:
–Original P(w | w ) c(w | w )
n n 1
n n 1

c(w ) n 1

–New P(w | w ) c(w | w ) 1

n n 1
n n 1

c(w ) V n 1

–But this change counts drastically:

• Too much weight given to unseen ngrams
• In practice, unsmoothed bigrams often work
better!
 Backoff Methods (e.g. Katz ’87)
• For e.g. a trigram model
– Compute unigram, bigram and trigram probabilities
– Usage:
• Where trigram unavailable back off to bigram if available, o.w. back
off to the current word’s unigram probability
• E.g An omnivorous unicorn
• NB: check errata pages for changes to figures
 Smoothing Summed Up

• Add-one smoothing (easy, but inaccurate)

– Add 1 to every word count (Note: this is type)
– Increment normalization factor by Vocabulary size: N (tokens)
+ V (types)
• Backoff models
– When a count for an n-gram is 0, back off to the count for the
(n-1)-gram
– These can be weighted – trigrams count more
 Summary
• N-gram probabilities can be used to estimate the
likelihood
– Of a word occurring in a context (N-1)
– Of a sentence occurring at all
• Smoothing techniques and backoff models deal with
problems of unseen words in corpus