Introduction to Information

Retrieval

Introduction to

Information Retrieval
CS276: Information Retrieval and
Web Search
Christopher Manning and Pandu
Nayak
Lecture 15: Learning to Rank

Introduction to Information
Retrieval

Machine learning for IR

ranking?

Sec. 15.4

Weve looked at methods for ranking documents in

IR
Cosine similarity, inverse document frequency,
proximity, pivoted document length normalization,
Pagerank,

Weve looked at methods for classifying documents

using supervised machine learning classifiers
Nave Bayes, Rocchio, kNN, SVMs

Surely we can also use machine learning to rank

the documents displayed in search results?
Sounds like a good idea
A.k.a. machine-learned relevance or learning to rank

Introduction to Information
Retrieval

Introduction to Information
Retrieval

Machine learning for IR ranking

This good idea has been actively
researched and actively deployed by
major web search engines in the last 7
or so years
Why didnt it happen earlier?
Modern supervised ML has been around for
about 20 years
Nave Bayes has been around for about 50
years

Introduction to Information
Retrieval

Machine learning for IR ranking

Theres some truth to the fact that the IR
community wasnt very connected to the ML
community
But there were a whole bunch of precursors:
Wong, S.K. et al. 1988. Linear structure in
information retrieval. SIGIR 1988.
Fuhr, N. 1992. Probabilistic methods in information
retrieval. Computer Journal.
Gey, F. C. 1994. Inferring probability of relevance
using the method of logistic regression. SIGIR 1994.
Herbrich, R. et al. 2000. Large Margin Rank
Boundaries for Ordinal Regression. Advances in
Large Margin Classifiers.

Introduction to Information
Retrieval

Why werent early attempts very

successful/influential?
Sometimes an idea just takes time to be
appreciated
Limited training data
Especially for real world use (as opposed to
writing academic papers), it was very hard to
gather test collection queries and relevance
judgments that are representative of real user
needs and judgments on documents returned
This has changed, both in academia and industry

Poor machine learning techniques

Insufficient customization to IR problem
Not enough features for ML to show value

Introduction to Information
Retrieval

Why wasnt ML much needed?

Traditional ranking functions in IR used a
very small number of features, e.g.,
Term frequency
Inverse document frequency
Document length

It was easy to tune weighting coefficients

by hand
And people did

Introduction to Information
Retrieval

Why is ML needed now?

Modern systems especially on the Web use a
great number of features:
Arbitrary useful features not a single unified model

Log frequency of query word in anchor text?

Query word in color on page?
# of images on page?
# of (out) links on page?
PageRank of page?
URL length?
URL contains ~?
Page edit recency?
Page length?

The New York Times (2008-06-03) quoted Amit

Singhal as saying Google was using over 200
such features.

Introduction to Information
Retrieval

Sec. 15.4.1

Simple example:
Using classification for ad hoc IR
Collect a training corpus of (q, d, r) triples
Relevance r is here binary (but may be multiclass,
with 37 values)

Document is represented by a feature vector

x = (, ) is cosine similarity, is minimum query
window size
is the the shortest text span that includes all
query words
Query term proximity is a very important new
weighting factor

Train a machine learning model to predict the

class r of a document-query pair

Introduction to Information
Retrieval

Sec. 15.4.1

Simple example:
Using classification for ad hoc IR
A linear score function is then
Score(d, q) = Score(, ) = a + b + c
And the linear classifier is
Decide relevant if Score(d, q) >
just like when we were doing text
classification

Introduction to Information
Retrieval

Sec. 15.4.1

Simple example:
Using classification for ad hoc IR
cosine score

0.05

R
R

0.025
R

R
R

R
R
R
N

N
N

R
N
N

0
2

R
N

4
5
Term proximity

Decision
surface

Introduction to Information
Retrieval

More complex example of using

classification for search ranking
[Nallapati 2004]

We can generalize this to classifier

functions over more features
We can use methods we have seen
previously for learning the linear classifier
weights

Introduction to Information
Retrieval

An SVM classifier for information

retrieval [Nallapati 2004]

Let g(r|d,q) = wf(d,q) + b

SVM training: want g(r|d,q) 1 for nonrelevant
documents and g(r|d,q) 1 for relevant
documents
SVM testing: decide relevant iff g(r|d,q) 0
Features are not word presence features (how would
you deal with query words not in your training data?)
but scores like the summed (log) tf of all query terms
Unbalanced data (which can result in trivial alwayssay-nonrelevant classifiers) is dealt with by
undersampling nonrelevant documents during
training (just take some at random)
[there are other
ways of doing this cf. Cao et al. later]

Introduction to Information
Retrieval

An SVM classifier for information

retrieval [Nallapati 2004]
Experiments:
4 TREC data sets
Comparisons with Lemur, a state-of-the-art
open source IR engine (Language Model (LM)based see IIR ch. 12)
Linear kernel normally best or almost as good
as quadratic kernel, and so used in reported
results
6 features, all variants of tf, idf, and tf.idf
scores

Introduction to Information
Retrieval

An SVM classifier for information

retrieval [Nallapati 2004]
Train \
Test
Disk 3

Disk 4-5

Disk 3

Disk 4-5

WT10G
(web)

LM

0.1785

0.2503

0.2666

SVM

0.1728

0.2432

0.2750

LM

0.1773

0.2516

0.2656

SVM

0.1646

0.2355

0.2675

At best the results are about equal to LM

Actually a little bit below

Papers advertisement: Easy to add more

features
This is illustrated on a homepage finding task on
WT10G:
Baseline LM 52% success@10, baseline SVM 58%
SVM with URL-depth, and in-link features: 78%
S@10

Introduction to Information
Retrieval

Sec. 15.4.2

Learning to rank
Classification probably isnt the right way to
think about approaching ad hoc IR:
Classification problems: Map to a unordered set of
classes
Regression problems: Map to a real value
Ordinal regression problems: Map to an ordered set
of classes
A fairly obscure sub-branch of statistics, but what we want
here

This formulation gives extra power:

Relations between relevance levels are modeled
Documents are good versus other documents for
query given collection; not an absolute scale of
goodness

Introduction to Information
Retrieval

Learning to rank
Assume a number of categories C of relevance exist
These are totally ordered: c1 < c2 < < cJ
This is the ordinal regression setup

Assume training data is available consisting of

document-query pairs represented as feature
vectors i and relevance ranking ci
We could do point-wise learning, where we try to
map items of a certain relevance rank to a
subinterval (e.g, Crammer et al. 2002 PRank)
But most work does pair-wise learning, where the
input is a pair of results for a query, and the class is
the relevance ordering relationship between them

Introduction to Information
Retrieval

Point-wise learning
Goal is to learn a threshold to separate
each rank

Introduction to Information
Retrieval

Sec. 15.4.2

Pairwise learning: The Ranking

SVM
[Herbrich et al. 1999, 2000; Joachims et al. 2002]

Aim is to classify instance pairs as

correctly ranked or incorrectly ranked
This turns an ordinal regression problem back
into a binary classification problem

We want a ranking function f such that

ci > ck iff f(i) > f(k)
or at least one that tries to do this with
minimal error
Suppose that f is a linear function
f(i) = wi

Introduction to Information
Retrieval

The Ranking SVM

[Herbrich et al. 1999, 2000; Joachims et al. 2002]

Ranking Model: f(i)

Sec. 15.4.2

Introduction to Information
Retrieval

Sec. 15.4.2

The Ranking SVM

[Herbrich et al. 1999, 2000; Joachims et al. 2002]

Then (combining the two equations on the

last slide):
ci > ck iff w(i k) > 0
Let us then create a new instance space
from such pairs:
u = (di, dj, q) = i k
zu = +1, 0, 1 as ci >,=,< ck
We can build model over just cases for
which zu = 1
From training data S = { u}, we train an
SVM

Introduction to Information
Retrieval

Two queries in the original

space

Introduction to Information
Retrieval

Two queries in the pairwise

space

Introduction to Information
Retrieval

Sec. 15.4.2

The Ranking SVM

[Herbrich et al. 1999, 2000; Joachims et al. 2002]

The SVM learning task is then like other

examples that we saw before
Find w and u 0 such that
wTw + C u is minimized, and
for all u such that zu < 0, wu 1 u

We can just do the negative zu, as ordering

is antisymmetric
You can again use SVMlight (or other good
SVM libraries) to train your model (SVMrank
specialization)

Introduction to Information
Retrieval

Aside: The SVM loss function

The minimization
minw wTw + C u
and for all u such that zu < 0, wu 1 u

can be rewritten as
minw (1/2C)wTw + u
and for all u such that zu < 0, u 1 (wu)

Now, taking = 1/2C, we can reformulate this

as
minw [1 (wu)]+ + wTw
Where []+ is the positive part (0 if a term is
negative)

Introduction to Information
Retrieval

Aside: The SVM loss function

Regularizer
Hinge loss
The reformulation
ofw
minw [1 (wu)]+ + wTw
shows that an SVM can be thought of as
having an empirical hinge loss
combined with a weight regularizer
Loss

wu

Introduction to Information
Retrieval

Adapting the Ranking SVM for

(successful) Information Retrieval
[Yunbo Cao, Jun Xu, Tie-Yan Liu, Hang Li, Yalou
Huang, Hsiao-Wuen Hon SIGIR 2006]

A Ranking SVM model already works well

Using things like vector space model scores as
features
As we shall see, it outperforms them in
evaluations

But it does not model important aspects of

practical IR well
This paper addresses two customizations
of the Ranking SVM to fit an IR utility
model

Introduction to Information
Retrieval

The ranking SVM fails to model the IR

problem well
1. Correctly ordering the most relevant
documents is crucial to the success of an IR
system, while misordering less relevant
results matters little
The ranking SVM considers all ordering violations
as the same

2. Some queries have many (somewhat)

relevant documents, and other queries few.
If we treat all pairs of results for a query
equally, queries with many results will
dominate the learning
But actually queries with few relevant results are at
least as important to do well on

Introduction to Information
Retrieval

Based on the LETOR test

collection

From Microsoft Research Asia

An openly available standard test collection with
pregenerated features, baselines, and research
results for learning to rank
Its availability has really driven research in this
area
OHSUMED, MEDLINE subcollection for IR

350,000 articles
106 queries
16,140 query-document pairs
3 class judgments: Definitely relevant (DR), Partially
Relevant (PR), Non-Relevant (NR)

TREC GOV collection (predecessor of GOV2, cf. IIR

p. 142)
1 million web pages
125 queries

Introduction to Information
Retrieval

Principal components projection of 2

queries
[solid = q12, open = q50; circle = DR, square = PR,
triangle = NR]

Introduction to Information
Retrieval

Ranking scale importance

discrepancy
[r3 = Definitely Relevant, r2 = Partially Relevant, r1 =
Nonrelevant]

Introduction to Information
Retrieval

Number of training documents per

query discrepancy [solid = q12, open =
q50]

Introduction to Information
Retrieval

Recap: Two Problems with Direct

Application of the Ranking SVM
Cost sensitiveness: negative effects of making errors
on top ranked documents
d: definitely relevant, p: partially relevant, n: not
relevant
ranking 1: p d p n n n n
ranking 2: d p n p n n n
Query normalization: number of instance pairs varies
according to query
q1: d p p n n n n
q2: d d p p p n n n n n
q1 pairs: 2*(d, p) + 4*(d, n) + 8*(p, n) = 14
q2 pairs: 6*(d, p) + 10*(d, n) + 15*(p, n) = 31

Introduction to Information
Retrieval

These problems are solved with a new

Loss function

weights for type of rank difference

Estimated empirically from effect on NDCG

weights for size of ranked result set

Linearly scaled versus biggest result set

Introduction to Information
Retrieval

Experiments
OHSUMED (from LETOR)
Features:
6 that represent versions of tf, idf, and tf.idf
factors
BM25 score (IIR sec. 11.4.3)
A scoring function derived from a probabilistic
approach to IR, which has traditionally done well in
TREC evaluations, etc.

Introduction to Information
Retrieval

Experimental Results
(OHSUMED)

Introduction to Information
Retrieval

MSN Search [now Bing]

Second experiment with MSN search
Collection of 2198 queries
6 relevance levels rated:

Definitive
Excellent
Good
Fair
Bad
Detrimental

8990
4403
3735
20463
36375
310

Introduction to Information
Retrieval

Experimental Results (MSN

search)

Introduction to Information
Retrieval

Alternative: Optimizing Rank-Based

Measures
[Yue et al. SIGIR 2007]

If we think that NDCG is a good

approximation of the users utility function
from a result ranking
Then, lets directly optimize this measure
As opposed to some proxy (weighted pairwise
prefs)

But, there are problems

Objective function no longer decomposes
Pairwise prefs decomposed into each pair

Objective function is flat or discontinuous

Introduction to Information
Retrieval

Discontinuity Example
NDCG computed using rank positions
Ranking via retrieval scores
Slight changes to model parameters
Slight changes to retrieval scores
No change to ranking
No change to NDCG

d1

d2

d3

NDCG = 0.63Retrieval Score 0.9 0.6 0.3

NDCG discontinuous
w.r.t model
parameters!

Rank

Relevance

Introduction to Information
Retrieval

Structural SVMs

[Tsochantaridis et

al., 2007]
Structural SVMs are a generalization of SVMs where the
output classification space is not binary or one of a set of
classes, but some complex object (such as a sequence
or a parse tree)
Here, it is a complete (weak) ranking of documents for a
query
The Structural SVM attempts to predict the complete
ranking for the input query and document set
The true labeling is a ranking where the relevant
documents are all ranked in the front, e.g.,
An incorrect labeling would be any other ranking, e.g.,

There are an intractable number of rankings, thus

an intractable number of constraints!

Introduction to Information
Retrieval

Structural SVM training

[Tsochantaridis et al., 2007]
Structural SVM training proceeds incrementally by starting with a
working set of constraints, and adding in the most violated
constraint at each iteration

Original SVM Problem

Exponential constraints
Most are dominated by a
small set of important
constraints

Structural SVM Approach

Repeatedly finds the next most

violated constraint
until a set of constraints which
is a good approximation is found

Introduction to Information
Retrieval

Other machine learning methods for

learning to rank
Of course!
Ive only presented the use of SVMs for
machine learned relevance, but other
machine learning methods have also been
used successfully

Boosting: RankBoost
Ordinal Regression loglinear models
Neural Nets: RankNet
(Gradient-boosted) Decisision Trees

Introduction to Information
Retrieval

The Limitation of Machine

Learning

Everything that we have looked at (and most

work in this area) produces linear models of
features by weighting different base features
This contrasts with most of the clever ideas of
traditional IR, which are nonlinear scalings and
combinations of basic measurements
log term frequency, idf, pivoted length
normalization

At present, ML is good at weighting features,

but not at coming up with nonlinear scalings
Designing the basic features that give good signals
for ranking remains the domain of human creativity

Introduction to Information
Retrieval

Summary
The idea of learning ranking functions has been
around for about 20 years
But only recently have ML knowledge, availability of
training datasets, a rich space of features, and
massive computation come together to make this a
hot research area
Its too early to give a definitive statement on what
methods are best in this area its still advancing
rapidly
But machine learned ranking over many features now
easily beats traditional hand-designed ranking
functions in comparative evaluations [in part by using the handdesigned functions as features!]

And there is every reason to think that the

importance of machine learning in IR will only
increase in the future.

Introduction to Information
Retrieval

Resources
IIR secs 6.1.23 and 15.4
LETOR benchmark datasets
Website with data, links to papers, benchmarks, etc.
http://research.microsoft.com/users/LETOR/
Everything you need to start research in this area!

Nallapati, R. Discriminative models for

information retrieval. SIGIR 2004.
Cao, Y., Xu, J. Liu, T.-Y., Li, H., Huang, Y. and
Hon, H.-W. Adapting Ranking SVM to
Document Retrieval, SIGIR 2006.
Y. Yue, T. Finley, F. Radlinski, T. Joachims. A
Support Vector Method for Optimizing Average
Precision. SIGIR 2007.