Introduction to Information Retrieval
Clustering: Definition
(Document) clustering is the process of grouping a set of
documents into clusters of similar documents.
Documents within a cluster should be similar.
Documents from different clusters should be dissimilar.
Clustering is the most common form of unsupervised
learning.
Unsupervised = there are no labeled or annotated data.
Introduction to Information Retrieval
Data set with clear cluster structure
Introduction to Information Retrieval
Classification vs. Clustering
Classification: supervised learning
Clustering: unsupervised learning
Classification: Classes are human-defined and part of the
input to the learning algorithm.
Clustering: Clusters are inferred from the data without
human input.
However, there are many ways of influencing the outcome of
clustering: number of clusters, similarity measure,
representation of documents, . . .
Introduction to Information Retrieval
The cluster hypothesis
Cluster hypothesis. Documents in the same cluster behave
similarly with respect to relevance to information needs. All
applications of clustering in IR are based (directly or indirectly) on
the cluster hypothesis.
Van Rijsbergens original wording: closely associated documents
tend to be relevant to the same requests.
Introduction to Information Retrieval
Yahoo! Hierarchy isnt clustering but is the kind
of output you want from clustering
www.yahoo.com/Science
(30)
agriculture
...
biology
physics
...
CS
...
space
...
...
dairy
botany
cell
AI
courses
crops
magnetism
HCI
agronomy
evolution
forestry
relativity
craft
missions
Introduction to Information Retrieval
Google News: automatic clustering gives an
effective news presentation metaphor
Introduction to Information Retrieval
Scatter/Gather: Cutting, Karger, and Pedersen
Sec. 16.1
Introduction to Information Retrieval
For visualizing a document collection and its
themes
Wise et al, Visualizing the non-visual PNNL
ThemeScapes, Cartia
[Mountain height = cluster size]
Introduction to Information Retrieval
Sec. 16.1
For improving search recall
Cluster hypothesis - Documents in the same cluster behave similarly
with respect to relevance to information needs
Therefore, to improve search recall:
Cluster docs in corpus a priori
When a query matches a doc D, also return other docs in the
cluster containing D
Hope if we do this: The query car will also return docs containing
automobile
Because clustering grouped together docs containing car with
those containing automobile.
Why might this happen?
Introduction to Information Retrieval
Desiderata for clustering
General goal: put related docs in the same cluster, put
unrelated docs in different clusters.
How do we formalize this?
The number of clusters should be appropriate for the data set
we are clustering.
Initially, we will assume the number of clusters K is given.
Later: Semiautomatic methods for determining K
Secondary goals in clustering
Avoid very small and very large clusters
Define clusters that are easy to explain to the user
Many others . . .
10
Introduction to Information Retrieval
Flat vs. Hierarchical clustering
Flat algorithms
Usually start with a random (partial) partitioning of docs into
groups
Refine iteratively
Main algorithm: K-means
Hierarchical algorithms
Create a hierarchy
Bottom-up, agglomerative
Top-down, divisive
11
Introduction to Information Retrieval
Hard vs. Soft clustering
Hard clustering: Each document belongs to exactly one
cluster.
More common and easier to do
Soft clustering: A document can belong to more than one
cluster.
Makes more sense for applications like creating browsable
hierarchies
You may want to put sneakers in two clusters:
sports apparel
shoes
You can only do that with a soft clustering approach.
12
Introduction to Information Retrieval
Flat algorithms
Flat algorithms compute a partition of N documents into a
set of K clusters.
Given: a set of documents and the number K
Find: a partition into K clusters that optimizes the chosen
partitioning criterion
Global optimization: exhaustively enumerate partitions,
pick optimal one
Not tractable
Effective heuristic method: K-means algorithm
13
Introduction to Information Retrieval
K-means
Perhaps the best known clustering algorithm
Simple, works well in many cases
Use as default / baseline for clustering documents
14
Introduction to Information Retrieval
Document representations in clustering
Vector space model
As in vector space classification, we measure relatedness
between vectors by Euclidean distance . . .
. . .which is almost equivalent to cosine similarity.
15
Introduction to Information Retrieval
K-means
Each cluster in K-means is defined by a centroid.
Objective/partitioning criterion: minimize the average
squared difference from the centroid
definition of centroid:
where we use to denote a cluster.
We try to find the minimum average squared difference by
iterating two steps:
reassignment: assign each vector to its closest centroid
recomputation: recompute each centroid as the average of
the vectors that were assigned to it in reassignment
16
Introduction to Information Retrieval
K-means algorithm
17
Introduction to Information Retrieval
Worked Example: Set of to be clustered
18
Introduction to Information Retrieval
Worked Example: Random selection of initial
centroids
Exercise: (i) Guess what the
optimal clustering into two clusters is in this case; (ii) compute the
centroids of the clusters
19
Introduction to Information Retrieval
Worked Example: Assign points to closest center
20
Introduction to Information Retrieval
Worked Example: Assignment
21
Introduction to Information Retrieval
Worked Example: Recompute cluster centroids
22
Introduction to Information Retrieval
Worked Example: Assign points to closest centroid
23
Introduction to Information Retrieval
Worked Example: Assignment
24
Introduction to Information Retrieval
Worked Example: Recompute cluster centroids
25
Introduction to Information Retrieval
Worked Example: Assign points to closest centroid
26
Introduction to Information Retrieval
Worked Example: Assignment
27
Introduction to Information Retrieval
Worked Example: Recompute cluster centroids
28
Introduction to Information Retrieval
Worked Example: Assign points to closest centroid
29
Introduction to Information Retrieval
Worked Example: Assignment
30
Introduction to Information Retrieval
Worked Example: Recompute cluster centroids
31
Introduction to Information Retrieval
Worked Example: Assign points to closest centroid
32
Introduction to Information Retrieval
Worked Example: Assignment
33
Introduction to Information Retrieval
Worked Example: Recompute cluster centroids
34
Introduction to Information Retrieval
Worked Example: Assign points to closest centroid
35
Introduction to Information Retrieval
Worked Example: Assignment
36
Introduction to Information Retrieval
Worked Example: Recompute cluster centroids
37
Introduction to Information Retrieval
Worked Example: Assign points to closest centroid
38
Introduction to Information Retrieval
Worked Example: Assignment
39
Introduction to Information Retrieval
Worked Example: Recompute cluster caentroids
40
Introduction to Information Retrieval
Worked Ex.: Centroids and assignments after
convergence
41
Introduction to Information Retrieval
K-means is guaranteed to converge
RSS = sum of all squared distances between document vector
and closest centroid
RSS decreases during each reassignment step.
because each vector is moved to a closer centroid
RSS decreases during each recomputation step.
There is only a finite number of clusterings.
Thus: We must reach a fixed point.
Assumption: Ties are broken consistently.
42
Introduction to Information Retrieval
Optimality of K-means
Convergence does not mean that we converge to the optimal
clustering!
This is the great weakness of K-means.
If we start with a bad set of seeds, the resulting clustering
can be horrible.
43
Introduction to Information Retrieval
Convergence Exercise: Suboptimal clustering
What is the optimal clustering for K = 2?
Do we converge on this clustering for arbitrary seeds di , dj?
44
Introduction to Information Retrieval
Initialization of K-means
Random seed selection is just one of many ways K-means
can be initialized.
Random seed selection is not very robust: Its easy to get a
suboptimal clustering.
Better ways of computing initial centroids:
Select seeds not randomly, but using some heuristic (e.g., filter
out outliers or find a set of seeds that has good coverage of
the document space)
Use hierarchical clustering to find good seeds
Select i (e.g., i = 10) different random sets of seeds, do a Kmeans clustering for each, select the clustering with lowest RSS
45
Introduction to Information Retrieval
Time complexity of K-means
Computing one distance of two vectors is O(M).
Reassignment step: O(KNM) (we need to compute KN
document-centroid distances)
Recomputation step: O(NM) (we need to add each of the
documents < M values to one of the centroids)
Assume number of iterations bounded by I
Overall complexity: O(IKNM) linear in all important
dimensions
46
Introduction to Information Retrieval
What is a good clustering?
Internal criteria
Example of an internal criterion: RSS in K-means
But an internal criterion often does not evaluate the actual
utility of a clustering in the application.
Alternative: External criteria
Evaluate with respect to a human-defined classification
47
Introduction to Information Retrieval
External criteria for clustering quality
Based on a gold standard data set, e.g., the Reuters
collection we also used for the evaluation of classification
Goal: Clustering should reproduce the classes in the gold
standard
(But we only want to reproduce how documents are divided
into groups, not the class labels.)
First measure for how well we were able to reproduce the
classes: purity
48
Introduction to Information Retrieval
External criterion: Purity
= {1, 2, . . . , K} is the set of clusters and
C = {c1, c2, . . . , cJ} is the set of classes.
For each cluster k : find class cj with most members nkj in k
Sum all nkj and divide by total number of points
49
Introduction to Information Retrieval
Sec. 16.3
Purity example
Cluster I
Cluster II
Cluster I: Purity = 1/6 (max(5, 1, 0)) = 5/6
Cluster II: Purity = 1/6 (max(1, 4, 1)) = 4/6
Cluster III: Purity = 1/5 (max(2, 0, 3)) = 3/5
Cluster III
Purity is (1/17)
(5 + 4 + 3) 0.71.
Introduction to Information Retrieval
Example for computing purity
Purity is (1/17) (5 + 4 + 3) 0.71.
51
Introduction to Information Retrieval
Rand index
Definition:
Based on 2x2 contingency table of all pairs of documents:
TP+FN+FP+TN is the total number of pairs.
There are
pairs for N documents.
Example:
= 136 in o//x example
Each pair is either positive or negative (the clustering puts
the two documents in the same or in different clusters) . . .
. . . and either true (correct) or false (incorrect): the
clustering decision is correct or incorrect.
52
Introduction to Information Retrieval
Rand Index: Example
As an example, we compute RI for the o//x example. We first
compute TP + FP. The three clusters contain 6, 6, and 5 points,
respectively, so the total number of positives or pairs of
documents that are in the same cluster is:
Of these, the x pairs in cluster 1, the o pairs in cluster 2, the
pairs in cluster 3, and the x pair in cluster 3 are true positives:
Thus, FP = 40 20 = 20. FN and TN are computed similarly.
53
Introduction to Information Retrieval
Rand measure for the o//x example
(20 + 72)/(20 + 20 + 24 + 72) 0.68.
54
Introduction to Information Retrieval
Two other external evaluation measures
Two other measures
Normalized mutual information (NMI)
How much information does the clustering contain about the
classification?
Singleton clusters (number of clusters = number of docs) have
maximum MI
Therefore: normalize by entropy of clusters and classes
F measure
Like Rand, but precision and recall can be weighted
55
Introduction to Information Retrieval
Normalized Mutual Information (NMI)
56
Introduction to Information Retrieval
Evaluation results for the o//x example
All four measures range from 0 (really bad clustering) to 1 (perfect
clustering).
57
Introduction to Information Retrieval
How many clusters?
Number of clusters K is given in many applications.
E.g., there may be an external constraint on K. Example: In the
case of Scatter-Gather, it was hard to show more than 1020
clusters on a monitor in the 90s.
What if there is no external constraint? Is there a right
number of clusters?
One way to go: define an optimization criterion
Given docs, find K for which the optimum is reached.
What optimiation criterion can we use?
We cant use RSS or average squared distance from centroid as
criterion: always chooses K = N clusters.
58
Introduction to Information Retrieval
Exercise
Your job is to develop the clustering algorithms for a
competitor to news.google.com
You want to use K-means clustering.
How would you determine K?
59
Introduction to Information Retrieval
Simple objective function for K (1)
Basic idea:
Start with 1 cluster (K = 1)
Keep adding clusters (= keep increasing K)
Add a penalty for each new cluster
Trade off cluster penalties against average squared distance
from centroid
Choose the value of K with the best tradeoff
60
Introduction to Information Retrieval
Simple objective function for K (2)
Given a clustering, define the cost for a document as
(squared) distance to centroid
Define total distortion RSS(K) as sum of all individual
document costs (corresponds to average distance)
Then: penalize each cluster with a cost
Thus for a clustering with K clusters, total cluster penalty is
K
Define the total cost of a clustering as distortion plus total
cluster penalty: RSS(K) + K
Select K that minimizes (RSS(K) + K)
Still need to determine good value for . . .
61
Introduction to Information Retrieval
Finding the knee in the curve
Pick the number of
clusters where curve
flattens. Here: 4 or 9.
62
Introduction to Information Retrieval
Resources
Chapter 16 of IIR
Resources at http://ifnlp.org/ir
K-means example
Keith van Rijsbergen on the cluster hypothesis (he was one of
the originators)
Bing/Carrot2/Clusty: search result clustering
63