1 information retrieval tanveer j siddiqui j k institute of applied physics & technology...
TRANSCRIPT
1
Information Retrieval
Tanveer J Siddiqui
J K Institute of Applied Physics & Technology
University of Allahabad
2
Lecture 1: Introduction
3
Information Retrieval
Information retrieval (IR) deals with the organization, storage, retrieval and evaluation of information relevant to user’s query.
A user having an information need formulates a request in the form of query written in natural language. The retrieval system responds by retrieving document that seems relevant to the query
4
Information Retrieval
Traditionally it has been accepted that information retrieval system does not return the actual information but the documents containing that information.
‘An information retrieval system does not inform (i.e. change the knowledge of) the user on the subject of her inquiry. It merely informs on the existence (or non-existence) and whereabouts of documents relating to her request.’
5
Information Retrieval Process
6
IR : Definition
IR is finding material of an unstructured nature (usually text) that satisfy an information need from within large collections
7
Unstructured data ?
- does not have a clear, semantically overt, easy-for-a computer structure.
8
IR vs. databases:Structured vs unstructured data
Structured data tends to refer to information in “tables”
Employee Manager Salary
Smith Jones 50000
Chang Smith 60000
50000Ivy Smith
Typically allows numerical range and exact match(for text) queries, e.g.,Salary < 60000 AND Manager = Smith.
9
Unstructured data
Typically refers to free text Allows
• Keyword queries including operators
• More sophisticated “concept” queries e.g.,• find all web pages dealing with drug abuse
Classic model for searching text documents
10
Semi-structured data
In fact almost no data is “unstructured” E.g., this slide has distinctly identified zones
such as the Title and Bullets Facilitates “semi-structured” search such as
• Title contains data AND Bullets contain search
… to say nothing of linguistic structure
11
More sophisticated semi-structured search
Title is about Object Oriented Programming AND Author something like stro*rup
where * is the wild-card operator The focus of XML search.
12
Unstructured (text) vs. structured (database) data in 1996
0
20
40
60
80
100
120
140
160
Data volume Market Cap
UnstructuredStructured
13
Unstructured (text) vs. structured (database) data in 2006
0
20
40
60
80
100
120
140
160
Data volume Market Cap
UnstructuredStructured
14
IR: An Example
Which plays of Shakespeare contain the words Brutus AND Caesar but NOT Calpurnia?
Simplest approach is to grep all of Shakespeare’s plays for Brutus and Caesar, then strip out lines containing Calpurnia?• Slow (for large corpora)
• NOT Calpurnia is non-trivial
• Other operations (e.g., find the word Romans near countrymen) not feasible
• Ranked retrieval (best documents to return)
15
How to avoid linear scanning ?
Index the documents in advance
16
Indexing
The process of transforming document text to some representation of it is known as indexing.
Different index structures might be used. One commonly used data structure by IR system is inverted index.
17
Information Retrieval Model
An IR model is a pattern that defines several aspects of retrieval procedure, for example, how the documents and user’s queries are represented how system retrieves relevant documents according to users’ queries & how retrieved documents are ranked.
18
IR Model
An IR model consists of
- a model for documents
- a model for queries and
- a matching function which compares queries to documents.
19
Classical IR Model
IR models can be classified as: Classical models of IR Non-Classical models of IR Alternative models of IR
20
Classical IR Model
based on mathematical knowledge that was easily recognized and well understood
simple, efficient and easy to implement The three classical information retrieval
models are: -Boolean -Vector and -Probabilistic models
21
Non-Classical models of IR
Non-classical information retrieval models are based on principles other than similarity, probability, Boolean operations etc. on which classical retrieval models are based on.
information logic model, situation theory model and interaction model.
22
Alternative IR models
Alternative models are enhancements of classical models making use of specific techniques from other fields.
Example:
Cluster model, fuzzy model and latent semantic indexing (LSI) models.
23
Information Retrieval Model
The actual text of the document and query is not used in the retrieval process. Instead, some representation of it.
Document representation is matched with query representation to perform retrieval
One frequently used method is to represent document as a set of index terms or keywords
24
Basics of Boolean IR model
Which plays of Shakespeare contain the words Brutus AND Caesar but NOT Calpurnia?
Document collection: A collection of Shakespeare's work
25
Binary Term-document matrix
1 if play contains word, 0 otherwise
Antony and Cleopatra Julius Caesar The Tempest Hamlet Othello Macbeth
Antony 1 1 0 0 0 1
Brutus 1 1 0 1 0 0
Caesar 1 1 0 1 1 1
Calpurnia 0 1 0 0 0 0
Cleopatra 1 0 0 0 0 0
mercy 1 0 1 1 1 1
worser 1 0 1 1 1 0
26
So we have a 0/1 vector for each term. To answer query: take the vectors for
Brutus, Caesar and Calpurnia (complemented) bitwise AND.
110100 AND 110111 AND 101111 = 100100.
27
Answers to query
Antony and Cleopatra, Act III, Scene ii
………… …………...
Hamlet, Act III, Scene ii ……………………. ……………………
28
Boolean retrieval model answers any query which is in the form of Boolean expression of terms.
29
Bigger corpora
Consider N = 1M documents, each with about 1K terms.
Avg 6 bytes/term incl spaces/punctuation • 6GB of data in the documents.
Say there are m = 500K distinct terms among these.
30
Can’t build the matrix
500K x 1M matrix has half-a-trillion 0’s and 1’s.
But it has no more than one billion 1’s.• matrix is extremely sparse.
What’s a better representation?• We only record the 1 positions.
Why?
31
Inverted index
For each term T, we must store a list of all documents that contain T.
Brutus
Calpurnia
Caesar
1 2 3 5 8 13 21 34
2 4 8 16 32 64128
13 16
What happens if the word Caesar is added to document 14?
we can use an array or a list.
32
Inverted index
Linked lists generally preferred to arrays• Dynamic space allocation
• Insertion of terms into documents easy
• Space overhead of pointers
Brutus
Calpurnia
Caesar
2 4 8 16 32 64 128
2 3 5 8 13 21 34
13 16
1
Dictionary Postings lists
Sorted by docID.
Posting
33
Inverted index construction
Tokenizer
Token stream. Friends Romans Countrymen
Linguistic modules
Modified tokens. friend roman countryman
Indexer
Inverted index.
friend
roman
countryman
2 4
2
13 16
1
More onthese later.
Documents tobe indexed.
34
Sequence of (Modified token, Document ID) pairs.
I did enact JuliusCaesar I was killed
i' the Capitol; Brutus killed me.
Doc 1
So let it be withCaesar. The noble
Brutus hath told youCaesar was ambitious
Doc 2
Term Doc #I 1did 1enact 1julius 1caesar 1I 1was 1killed 1i' 1the 1capitol 1brutus 1killed 1me 1so 2let 2it 2be 2with 2caesar 2the 2noble 2brutus 2hath 2told 2you 2
caesar 2was 2ambitious 2
Indexer steps
35
Sort by terms(Core indexing step.). Term Doc #ambitious 2be 2brutus 1brutus 2capitol 1caesar 1caesar 2caesar 2did 1enact 1hath 1I 1I 1i' 1it 2julius 1killed 1killed 1let 2me 1noble 2so 2the 1the 2told 2you 2was 1was 2with 2
Term Doc #I 1did 1enact 1julius 1caesar 1I 1was 1killed 1i' 1the 1capitol 1brutus 1killed 1me 1so 2let 2it 2be 2with 2caesar 2the 2noble 2brutus 2hath 2told 2you 2caesar 2was 2ambitious 2
36
Multiple term entries in a single document are merged.
Frequency information is added.
Term Doc # Term freqambitious 2 1be 2 1brutus 1 1brutus 2 1capitol 1 1caesar 1 1caesar 2 2did 1 1enact 1 1hath 2 1I 1 2i' 1 1it 2 1julius 1 1killed 1 2let 2 1me 1 1noble 2 1so 2 1the 1 1the 2 1told 2 1you 2 1was 1 1was 2 1with 2 1
Term Doc #ambitious 2be 2brutus 1brutus 2capitol 1caesar 1caesar 2caesar 2did 1enact 1hath 1I 1I 1i' 1it 2julius 1killed 1killed 1let 2me 1noble 2so 2the 1the 2told 2you 2was 1was 2with 2
37
The result is split into a Dictionary file and a Postings file.
Doc # Freq2 12 11 12 11 11 12 21 11 12 11 21 12 11 11 22 11 12 12 11 12 12 12 11 12 12 1
Term N docs Coll freqambitious 1 1be 1 1brutus 2 2capitol 1 1caesar 2 3did 1 1enact 1 1hath 1 1I 1 2i' 1 1it 1 1julius 1 1killed 1 2let 1 1me 1 1noble 1 1so 1 1the 2 2told 1 1you 1 1was 2 2with 1 1
Term Doc # Freqambitious 2 1be 2 1brutus 1 1brutus 2 1capitol 1 1caesar 1 1caesar 2 2did 1 1enact 1 1hath 2 1I 1 2i' 1 1it 2 1julius 1 1killed 1 2let 2 1me 1 1noble 2 1so 2 1the 1 1the 2 1told 2 1you 2 1was 1 1was 2 1with 2 1
38
Doc # Freq
2 12 11 12 11 11 12 21 11 12 11 21 12 11 11 22 11 12 12 11 12 12 12 11 12 12 1
Term N docs Coll freqambitious 1 1be 1 1brutus 2 2capitol 1 1caesar 2 3did 1 1enact 1 1hath 1 1I 1 2i' 1 1it 1 1julius 1 1killed 1 2let 1 1me 1 1noble 1 1so 1 1the 2 2told 1 1you 1 1was 2 2with 1 1
Pointers
Terms
39
The index we just built
How do we process a query?
40
Query processing: AND
Consider processing the query:Brutus AND Caesar
• Locate Brutus in the Dictionary;• Retrieve its postings.
• Locate Caesar in the Dictionary;• Retrieve its postings.
• “Merge” the two postings:128
34
2 4 8 16 32 64
1 2 3 5 8 13
21
Brutus
Caesar
41
34
1282 4 8 16 32 64
1 2 3 5 8 13 21
The merge
Walk through the two postings simultaneously, in time linear in the total number of postings entries
128
34
2 4 8 16 32 64
1 2 3 5 8 13 21
Brutus
Caesar2 8
If the list lengths are x and y, the merge takes O(x+y)operations.Crucial: postings sorted by docID.
42
Merging Algorithm
Merge(p,q)
1 Start
2. Ans ()
3. While p<> nil and q <> nil do
if pdocID = q docID
then ADD(answer, pdocID) // add to result and advance pointers
else if pdocID < qdocID
then p pnext
else q qnext
4. end {of algo}
43
Boolean queries: Exact match The Boolean Retrieval model is being able to
ask a query that is a Boolean expression:• Boolean Queries are queries using AND, OR and
NOT to join query terms
• Views each document as a set of words
• Is precise: document matches condition or not.
Primary commercial retrieval tool for 3 decades.
Professional searchers (e.g., lawyers) still like Boolean queries.
44
Example: WestLaw http://www.westlaw.com/
Largest commercial (paying subscribers) legal search service (started 1975; ranking added 1992)
Tens of terabytes of data; 700,000 users Majority of users still use boolean queries
45
Merging: More general merges
Consider an arbitrary Boolean formula:
(Brutus OR Caesar) AND NOT
(Antony OR Cleopatra)
46
Query optimization
What is the best order for query processing? Consider a query that is an AND of t terms. For each of the t terms, get its postings, then AND
them together.
Brutus
Calpurnia
Caesar
1 2 3 5 8 16 21 34
2 4 8 16 32 64128
13 16
Query: Brutus AND Calpurnia AND Caesar
47
Query Optimization
How to organize the work of getting results for a query so that the amount of work is reduced.
48
Query optimization example
Process in order of increasing freq:• start with smallest set, then keep cutting
further.
Brutus
Calpurnia
Caesar
1 2 3 5 8 13 21 34
2 4 8 16 32 64128
13 16
This is why we keptfreq in dictionary
Execute the query as (Caesar AND Brutus) AND Calpurnia.
49
More general optimization
e.g., (madding OR crowd) AND (ignoble OR strife)
Get freq’s for all terms.Estimate the size of each OR by the
sum of its freq’s (conservative).Process in increasing order of OR
sizes.
50
Beyond term search
Phrases?• Allahabad University
Proximity: Find Murty NEAR Infosys.• Need index to capture position information in
docs.
Find documents with (author = Zufrasky) AND (text contains Retrieval).
51
What else to consider ?
1 vs. 0 occurrence of a search term• 2 vs. 1 occurrence
• 3 vs. 2 occurrences, etc.
• Usually more seems better
Need term frequency information in docs
52
Ranking search results
Boolean queries give inclusion or exclusion of docs.
Requires precise language for building query expressions ( instead of free text )
Often we want to rank/group results
53
Clustering and classification
Given a set of docs, group them into clusters based on their contents.
Given a set of topics, plus a new doc D, decide which topic(s) D belongs to.
54
The web and its challenges
Unusual and diverse documentsUnusual and diverse users, queries, information
needsBeyond terms, exploit ideas from social
networks• link analysis, clickstreams ...
How do search engines work? And how can we make them better?
55
More sophisticated information retrieval
Cross-language information retrieval Question answering Summarization Text mining …
56
57
58
The computational cost involved in adopting a full text logical view (i.e. using full set of words to represent a document) is high.
Hence, some text operations are usually performed to reduce the set of representative keywords.
59
Two most commonly used text operations are:
1. Stop word elimination and
2. Stemming Zipf’s law
60
Stop word elimination involves removal of grammatical or function words while stemming reduces distinct words to their common grammatical root
61
Indexing
Most of the indexing techniques involve identifying good document descriptors, such as keywords or terms, to describe information content of the documents.
A good descriptor is one that helps in describing the content of the document and in discriminating the document from other documents in the collection.
62
Term can be a single word or it can be multi-word phrases.
Example: Design Features of Information
Retrieval systems can be represented by the set of terms : Design, Features, Information,
Retrieval, systemsor by the set of terms: Design, Features, Information
Retrieval, Information Retrieval systems
63
Luhn’s early Assumption
Luhn assumed that frequency of word occurrence in an article gives meaningful identification of their content.
discrimination power for index terms is a function of the rank order of their frequency of occurrence
64
Stop Word Elimination
Stop words are high frequency words, which have little semantic weight and are thus unlikely to help with retrieval.
Such words are commonly used in documents, regardless of topics; and have no topical specificity.
65
Example :
articles (“a”, “an” “the”) and prepositions (e.g. “in”, “of”, “for”, “at” etc.).
66
Stop Word Elimination
Advantage Eliminating these words can result in
considerable reduction in number of index terms without losing any significant information.
Disadvantage It can sometimes result in elimination of terms
useful for searching, for instance the stop word A in Vitamin A.
Some phrases like “to be or not to be” consist entirely of stop words.
67
Stop Words
About, above, accordingly, afterwards, again, against, alone, along, already, am, among, amongst, and, another, any, anyone, anything, anywhere, around, as, aside, awfully, be, because
68
Stemming
Stemming normalizes morphological variants
It removes suffixes from the words to reduce them to some root form e.g. the words compute, computing, computes and computer will all be reduced to same word stem comput.
69
Stemming
Porter Stemmer(1980). Example:
The stemmed representation of
Design Features of Information Retrieval systems
will be
{design, featur, inform, retriev, system}
70
Stemming
stemming throws away useful distinction. In some cases it may be useful to help conflate similar terms resulting in increased recall in others it may be harmful resulting in reduced precision
71
Zipf’s law
Zipf law
frequency of words multiplied by their ranks in a large corpus is approximately constant, i.e.
constant rank frequency
72
Relationship between frequency of words and its rank order
73
Zipf law
f
74
75
Simple indexing Scheme based on Zipf’s law
76
Luhn’s assumptions
Luhn(1958) attempted to quantify the discriminating power of the terms by associating their frequency of occurrence (term frequency) within the document. He postulated that:
-The high frequency words are quite common (function words)
- low frequency words are rare words - medium frequency words are useful for
indexing
77
Boolean model
the oldest of the three classical models. is based on Boolean logic and classical
set theory. represents documents as a set of
keywords, usually stored in an inverted file.
78
Boolean model
Users are required to express their queries as a boolean expression consisting of keywords connected with boolean logical operators (AND, OR, NOT).
Retrieval is performed based on whether or not document contains the query terms.
79
Boolean model
Given a finite set
T = {t1, t2, ...,ti,...,tm}
of index terms, a finite set
D = {d1, d2, ...,dj,...,dn}
of documents and a boolean expression in a normal form - representing a query Q as follows:
},{θi),θ ( iii ttQ
80
Boolean model
1. The set Ri of documents are obtained that contain or not term ti:
Ri = { },
where
2. Set operations are used to retrieve documents in response to Q:
jj did | ,},{ ii tti
jiji dtdt means
iR