Introduction to Information Retrieval

Information Retrieval and Data Mining (AT71.07) Comp. Sc. and Inf. Mgmt. Asian Institute of Technology
Instructor: Dr. Sumanta Guha Slide Sources: Introduction to Information Retrieval book slides from Stanford University, adapted and supplemented Chapter 2: The term vocabulary and postings lists
1

Introduction to Information Retrieval

Introduction to

Information Retrieval
CS276: Information Retrieval and Web Search Christopher Manning and Prabhakar Raghavan Lecture 2: The term vocabulary and postings lists

Introduction to Information Retrieval

Ch. 1

Recap of the previous lecture

Basic inverted indexes:
Structure: Dictionary and Postings

Key step in construction: Sorting

Boolean query processing

Intersection by linear time merging Simple optimizations

Overview of course topics

Introduction to Information Retrieval

Plan for this lecture

Elaborate basic indexing Preprocessing to form the term vocabulary
Documents Tokenization What terms do we put in the index?

Postings
Faster merges: skip lists Positional postings and phrase queries

Introduction to Information Retrieval

Recall the basic indexing pipeline

Documents to be indexed.

Friends, Romans, countrymen. Tokenizer

Token stream.

Friends Romans
Linguistic modules

Countrymen

Modified tokens. = terms Inverted index.

friend
Indexer friend
roman

roman

countryman

2
1

4
2 16

countryman

13

Introduction to Information Retrieval

Sec. 2.1

Parsing a document
What format is it in?
pdf/word/excel/html?

What language is it in? What character set is in use?

Each of these is a classification problem, which we will study later in the course. But these tasks are often done heuristically

Introduction to Information Retrieval

Sec. 2.1

Complications: Format/language
Documents being indexed can include docs from many different languages
A single index may have to contain terms of several languages.

Sometimes a document or its components can contain multiple languages/formats

French email with a German pdf attachment.

What is a unit document?

A file? An email? (Perhaps one of many in an mbox.) An email with 5 attachments? A group of files (PPT or LaTeX as HTML pages)

Introduction to Information Retrieval

TOKENS AND TERMS

Introduction to Information Retrieval

Definitions
Word A delimited string of characters as it appears in the text. Term A normalized word (case, morphology, spelling etc); an equivalence class of words. Token An instance of a word or term occurring in a document. Type The same as a term in most cases: an equivalence class of tokens.

Introduction to Information Retrieval

Sec. 2.2.1

Tokenization
Input: Friends, Romans and Countrymen Output: Tokens
Friends Romans and Countrymen

A token is an instance of a sequence of characters Each such token is now a candidate for an index entry, after further processing
Described below

But what are valid tokens to emit?

Introduction to Information Retrieval

Sec. 2.2.1

Tokenization
Issues in tokenization: Finlands capital
Finland? Finlands? Finlands? Hewlett-Packard Hewlett and Packard as two tokens?
state-of-the-art: break up hyphenated sequence. co-education lowercase, lower-case, lower case ?
It can be effective to get the user to put in possible hyphens

San Francisco: one token or two?

How do you decide it is one token?

Introduction to Information Retrieval

Sec. 2.2.1

Numbers
3/20/91 Mar. 12, 1991 20/3/91 55 B.C. B-52 My PGP key is 324a3df234cb23e (800) 234-2333 Often have embedded spaces Older IR systems may not index numbers
But often very useful: think about things like looking up error codes/stacktraces on the web (One answer is using n-grams: Lecture 3)

Will often index meta-data separately

Creation date, format, etc.

Introduction to Information Retrieval

Sec. 2.2.1

Tokenization: language issues

French
L'ensemble one token or two?
L ? L ? Le ? Want lensemble to match with un ensemble Until at least 2003, it didnt on Google Internationalization!

German noun compounds are not segmented

Lebensversicherungsgesellschaftsangestellter life insurance company employee German retrieval systems benefit greatly from a compound splitter module
Can give a 15% performance boost for German

Introduction to Information Retrieval

Sec. 2.2.1

Tokenization: language issues

Chinese, Japanese and Thai have no spaces between words:
Not always guaranteed a unique tokenization

Further complicated in Japanese, with multiple alphabets intermingled

Dates/amounts in multiple formats
500$500K(6,000)

Katakana

Hiragana

Kanji

Romaji

End-user can express query entirely in hiragana!

Introduction to Information Retrieval

Sec. 2.2.1

Tokenization: language issues

Arabic (or Hebrew) is basically written right to left, but with certain items like numbers written left to right Words are separated, but letter forms within a word form complex ligatures start Algeria achieved its independence in 1962 after 132 years of French occupation.
With Unicode, the surface presentation is complex, but the stored form is straightforward

Introduction to Information Retrieval

Sec. 2.2.2

Stop words
With a stop list, you exclude from the dictionary entirely the commonest words. Intuition:
They have little semantic content: the, a, and, to, be There are a lot of them: ~30% of postings for top 30 words

But the trend is away from doing this:

Good compression techniques (lecture 5) means the space for including stopwords in a system is very small Good query optimization techniques (lecture 7) mean you pay little at query time for including stop words. You need them for:
Phrase queries: King of Denmark Various song titles, etc.: Let it be, To be or not to be Relational queries: flights to London

Introduction to Information Retrieval

Sec. 2.2.3

Normalization to terms
We need to normalize words in indexed text as well as query words into the same form
We want to match U.S.A. and USA

Result is terms: a term is a (normalized) word type, which is an entry in our IR system dictionary We most commonly implicitly define equivalence classes of terms by, e.g.,
deleting periods to form a term
U.S.A., USA USA

deleting hyphens to form a term

anti-discriminatory, antidiscriminatory antidiscriminatory

Introduction to Information Retrieval

Sec. 2.2.3

Normalization: other languages

Accents: e.g., French rsum vs. resume. Umlauts: e.g., German: Tuebingen vs. Tbingen
Should be equivalent

Most important criterion:

How are your users like to write their queries for these words?

Even in languages that standardly have accents, users often may not type them
Often best to normalize to a de-accented term
Tuebingen, Tbingen, Tubingen Tubingen

Introduction to Information Retrieval

Sec. 2.2.3

Normalization: other languages

Normalization of things like date forms
730 vs. 7/30 Japanese use of kana vs. Chinese characters

Tokenization and normalization may depend on the language and so is intertwined with language detection
Morgen will ich in MIT

Is this German mit?

Crucial: Need to normalize indexed text as well as query terms into the same form

Introduction to Information Retrieval

Sec. 2.2.3

Case folding
Reduce all letters to lower case
exception: upper case in mid-sentence?
e.g., General Motors Fed vs. fed SAIL vs. sail

Often best to lower case everything, since users will use lowercase regardless of correct capitalization

Google example:
Query C.A.T. #1 result is for cat (well, Lolcats) not Caterpillar Inc.

Introduction to Information Retrieval

Sec. 2.2.3

Normalization to terms
An alternative to equivalence classing is to do asymmetric query expansion An example of where this may be useful
Enter: window Enter: windows Enter: Windows Search: window, windows Search: Windows, windows, window Search: Windows

Potentially more powerful, but less efficient

Introduction to Information Retrieval

Thesauri and soundex

Do we handle synonyms and homonyms?
E.g., by hand-constructed equivalence classes
car = automobile color = colour

We can rewrite to form equivalence-class terms

When the document contains automobile, index it under carautomobile (and vice-versa)

Or we can expand a query

When the query contains automobile, look under car as well

What about spelling mistakes?

One approach is soundex, which forms equivalence classes of words based on phonetic heuristics (Muller = Mueller, color = colour)

More in lectures 3 and 9

Introduction to Information Retrieval

Sec. 2.2.4

Lemmatization
Reduce inflectional/variant forms to base form E.g.,
am, are, is be car, cars, car's, cars' car

the boy's cars are different colors the boy car be different color
Lemmatization implies doing proper reduction to dictionary headword form

Introduction to Information Retrieval

Sec. 2.2.4

Stemming
Reduce terms to their roots before indexing Stemming suggest crude affix chopping
language dependent e.g., automate(s), automatic, automation all reduced to automat.

for example compressed and compression are both accepted as equivalent to compress.

for exampl compress and compress ar both accept as equival to compress

Introduction to Information Retrieval

Sec. 2.2.4

Porters algorithm
Commonest algorithm for stemming English
Results suggest its at least as good as other stemming options

Conventions + 5 phases of reductions

phases applied sequentially each phase consists of a set of commands sample convention: Of the rules in a compound command, select the one that applies to the longest suffix.

Introduction to Information Retrieval

Sec. 2.2.4

Typical rules in Porter

sses ss ies i ational ate tional tion

Weight of word sensitive rules (m>1) EMENT

replacement replac cement cement

Introduction to Information Retrieval

Sec. 2.2.4

Other stemmers
Other stemmers exist, e.g., Lovins stemmer
http://www.comp.lancs.ac.uk/computing/research/stemming/general/lovins.htm

Single-pass, longest suffix removal (about 250 rules)

Full morphological analysis at most modest benefits for retrieval Do stemming and other normalizations help?
English: very mixed results. Helps recall for some queries but harms precision on others
E.g., operative (dentistry) oper

Definitely useful for Spanish, German, Finnish,

30% performance gains for Finnish!

Introduction to Information Retrieval

Sec. 2.2.4

Language-specificity
Many of the above features embody transformations that are
Language-specific and Often, application-specific

These are plug-in addenda to the indexing process Both open source and commercial plug-ins are available for handling these

Introduction to Information Retrieval

Sec. 2.2

Dictionary entries first cut

ensemble.french

.japanese
MIT.english

mit.german
guaranteed.english

entries.english
sometimes.english tokenization.english

These may be grouped by language (or not). More on this in ranking/query processing.

Introduction to Information Retrieval

Sec. 1.3

Stemming and Lemmatization

a. b. c. d. Exercise 2.1: Are the following statements true or false: In a Boolean retrieval system, stemming never lowers precision. In a Boolean retrieval system, stemming never lowers recall. Stemming increases the size of the vocabulary. Stemming should be invoked at index time, but not while processing a query. Exercise 2.2: Suggest what normalized form should be used for these words (including the word itself as a possibility): Cos Shiite contd Hawaii ORourke
30

a. b. c. d. e.

Introduction to Information Retrieval

Sec. 1.3

Stemming and Lemmatization

Exercise 2.3: The following pairs of words are stemmed to the same form by the Porter stemmer. Which pairs, would you argue, should not be conflated? Give your reasoning. a. abandon/abandonment b. absorbency/absorbent c. marketing/markets d. university/universe e. volume/volumes
31

Introduction to Information Retrieval

Sec. 1.3

Stemming and Lemmatization

Porters stemming algorithm, phase 1 rules (select rule which applies to longest suffix) Rule Example SSES SS caresses caress IES I ponies poni SS SS caress caress S cats cat Exercise 2.4: For the Porter stemmer rule group shown above: a. What is the purpose of including an identity rule SS SS b. Applying just this rule group, what will the following words be stemmed to? circus canaries boss c. What rule should be added to correctly stem pony? d. The stemming for ponies and pony might seem strange. Does it have a 32 deleterious effect on retrieval? Why or why not?

Introduction to Information Retrieval

FASTER POSTINGS MERGES: SKIP POINTERS/SKIP LISTS

Introduction to Information Retrieval

Sec. 2.3

Recall basic merge

Walk through the two postings simultaneously, in time linear in the total number of postings entries

2
2 8 1

4
2

8
3

41
8

48
11

64
17

128
21

Brutus

31 Caesar

If the list lengths are m and n, the merge takes O(m+n) operations.
Can we do better? Yes (if index isnt changing too fast).

Introduction to Information Retrieval

Augment postings with skip pointers (at indexing time)

41 128

Sec. 2.3

2
11

41

48
31

64

128

11

17

21

31

Why? To skip postings that will not figure in the search results. How? Where do we place skip pointers?

Introduction to Information Retrieval

Sec. 2.3

Query processing with skip pointers

41 128

2
11

41

48
31

64

128

11

17

21

31

Suppose weve stepped through the lists until we process 8 on each list. We match it and advance. We then have 41 and 11 on the lower. 11 is smaller. But the skip successor of 11 on the lower list is 31, so we can skip ahead past the intervening postings.

Introduction to Information Retrieval IntersectWithSkips(p1, p2) 1 answer <>

2 while p1 NIL and p2 NIL

3 do if docID(p1) = docID(p2) 4 5 then ADD(answer, docID(p1)) p1 next(p1)

6
7 8 9

p2 next(p2)
else if docID(p1) < docID(p2) then if hasSkip(p1) and (docID(skip(p1)) docID(p2)) then while hasSkip(p1) and (docID(skip(p1)) docID(p2))

10
11 12 13

do p1 skip(p1)
else p1 next(p1) else if hasSkip(p2) and (docID(skip(p2)) docID(p1)) then while hasSkip(p2) and (docID(skip(p2)) docID(p1))

14
15 16 return answer

do p2 skip(p2)
else p2 next(p2)
37

Introduction to Information Retrieval

Sec. 2.3

Where do we place skips?

Tradeoff:
More skips shorter skip spans more likely to skip. But lots of comparisons to skip pointers. Fewer skips few pointer comparison, but then long skip spans few successful skips.

Introduction to Information Retrieval

Sec. 2.3

Placing skips
Simple heuristic: for postings of length L, use L evenly-spaced skip pointers. This ignores the distribution of query terms. Easy if the index is relatively static; harder if L keeps changing because of updates. This definitely used to help; with modern hardware it may not (Bahle et al. 2002) unless youre memorybased
The I/O cost of loading a bigger postings list can outweigh the gains from quicker in memory merging!

Introduction to Information Retrieval

Skips
Exercise 2.5: Why are skip pointers not useful for queries of the form x OR y? Exercise 2.6: We have a two-word query. For one term the postings list consists of the 16 entries [4, 6, 10, 12, 14, 16, 18, 20, 22, 32, 47, 81, 120, 122, 157, 180] and for the other it consists of only 1 entry [47] Work out how many comparisons would be done to intersect the two postings lists with the following strategies. a. Using standard postings lists. b. Using postings lists with skip pointers, with the suggested skip length of P.

40

Introduction to Information Retrieval

Skips
Exercise 2.7: Consider a postings intersection between this postings list with skip pointers:

3 5 9 15 24 39 60 68 75 81 84 89 92 96 97 100 115 and the following intermediate result posting list (which hence has no skip pointers): 3 5 89 95 97 99 100 101 Trace through the postings intersection algorithm. a. How often is a skip pointer followed (i.e., p1 advanced to skip(p1))? b. How many postings comparisons will be made by this algorithm while intersecting the two lists? c. How many comparisons would be made if the postings lists are intersected without the use of skip pointers? 41

Introduction to Information Retrieval

PHRASE QUERIES AND POSITIONAL INDEXES

Introduction to Information Retrieval

Sec. 2.4

Phrase queries

Want to be able to answer queries such as stanford university as a phrase Thus the sentence I went to university at Stanford is not a match. The sentence The inventor Stanford Ovshinsky never went to university is not a match.
The concept of phrase queries has proven easily understood by users; one of the few advanced search ideas that works Many more queries are implicit phrase queries

For this, it no longer suffices to store only <term : docs> entries

Introduction to Information Retrieval

Sec. 2.4.1

A first attempt: Biword indexes

Index every consecutive pair of terms in the text as a phrase For example the text Friends, Romans, Countrymen would generate the biwords
friends romans romans countrymen

Each of these biwords is now a dictionary term Two-word phrase query-processing is now immediate.

Introduction to Information Retrieval

Sec. 2.4.1

Longer phrase queries

Longer phrases are processed as we did with wildcards: stanford university palo alto can be broken into the Boolean query on biwords: stanford university AND university palo AND palo alto Without the docs, we cannot verify that the docs matching the above Boolean query do contain the phrase.
Can have false positives!

Introduction to Information Retrieval

Sec. 2.4.1

Extended biwords
Parse the indexed text and perform part-of-speech-tagging (POST). Bucket the terms into (say) Nouns (N) and articles/prepositions (X). Call any string of terms of the form NX*N an extended biword. Each such extended biword is now made a term in the dictionary. Example: catcher in the rye N X X N Query processing: parse it into Ns and Xs Segment query into enhanced biwords Look up in index: catcher rye

Introduction to Information Retrieval

Sec. 2.4.1

Issues for biword indexes

False positives, as noted before Index blowup due to bigger dictionary
Infeasible for more than biwords, big even for them

Biword indexes are not the standard solution (for all biwords) but can be part of a compound strategy

Introduction to Information Retrieval

Sec. 2.4.2

Solution 2: Positional indexes

In the postings, store, for each term the position(s) in which tokens of it appear:
<term, number of docs containing term; doc1: frequency of the term; position1, position2 ; doc2: frequency of the term; position1, position2 ; etc.>

Introduction to Information Retrieval

Sec. 2.4.2

Positional index example

<be: 993427; 1: 6; 7, 18, 33, 72, 86, 231; 2: 2; 3, 149; 4: 5; 17, 191, 291, 430, 434; 5: 11; 363, 367, >

Which of docs 1,2,4,5 could contain to be or not to be?

For phrase queries, we use a merge algorithm recursively at the document level But we now need to deal with more than just equality

Introduction to Information Retrieval

Sec. 2.4.2

Processing a phrase query

Extract inverted index entries for each distinct term: to, be, or, not. Merge their doc:position lists to enumerate all positions with to be or not to be.
to: 2: 5; 1,17,74,222,551; 4: 5; 8,16,190,429,433; 7: 3; 13,23,191; ... be: 1: 2;17,19; 4: 5;17,191,291,430,434; 5: 3; 14,19,101; ...

Same general method for proximity searches

Introduction to Information Retrieval

Sec. 2.4.2

Proximity queries
LIMIT! /3 STATUTE /3 FEDERAL /2 TORT
Again, here, /k means within k words of.

Clearly, positional indexes can be used for such queries; biword indexes cannot. Exercise: Adapt the linear merge of postings to handle proximity queries. Can you make it work for any value of k?
This is a little tricky to do correctly and efficiently See Figure 2.12 of IIR Theres likely to be a problem on it!

Introduction to Information Retrieval PositionalIntersect(p1, p2, k) 1 answer <> l is a moving window of positions 2 while p1 nil and p2 nil of the second word in the current 3 do if docID(p1) = docID(p2) doc which are within k of the 4 then l <> current position of the first word. 5 pp1 positions(p1) 6 pp2 positions(p2) 7 while pp1 nil 8 do while pp2 nil 9 do if |pos(pp1) pos(pp2)| k 10 then ADD(l , pos(pp2)) 11 else if pos(pp2) > pos(pp1) 12 then break 13 pp2 next(pp2) 14 while l <> and |l[0] pos(pp1)| > k 15 do DELETE(l[0]) 16 for each ps l 17 do ADD(answer, <docID(p1), pos(pp1), ps>) 18 pp1 next(pp1) For each successive position of the 19 p1 next(p1) first word: moves the head of the 20 p2 next(p2) window l up till at most k away from 21 else if docID(p1) < docID(p2) the new position of the first word; 22 then p1 next(p1) deletes tail of the window till within 23 else p2 next(p2) k of the first word. 24 return answer

52

Introduction to Information Retrieval

Sec. 2.4.2

Positional index size

You can compress position values/offsets: well talk about that in lecture 5 Nevertheless, a positional index expands postings storage substantially Nevertheless, a positional index is now standardly used because of the power and usefulness of phrase and proximity queries whether used explicitly or implicitly in a ranking retrieval system.

Introduction to Information Retrieval

Sec. 2.4.2

Positional index size

Need an entry for each occurrence, not just once per document Index size depends on average document size Why?
Average web page has <1000 terms SEC filings, books, even some epic poems easily 100,000 terms

Consider a term with frequency 0.1%

Document size
1000 100,000 Postings 1 1
Positional postings

1 100

Introduction to Information Retrieval

Sec. 2.4.2

Rules of thumb
A positional index is 24 as large as a non-positional index Positional index size 3550% of volume of original text Caveat: all of this holds for English-like languages

Introduction to Information Retrieval

Sec. 2.4.3

Combination schemes
These two approaches can be profitably combined
For particular phrases (Michael Jackson, Britney Spears) it is inefficient to keep on merging positional postings lists
Even more so for phrases like The Who

Williams et al. (2004) evaluate a more sophisticated mixed indexing scheme

A typical web query mixture was executed in of the time of using just a positional index It required 26% more space than having a positional index alone

Introduction to Information Retrieval

Positional Postings and Phrase Queries

Exercise 2.8: Assume a biword index. Give an example of a document that will be returned for a query of New York University but is actually a false positive.

57

Introduction to Information Retrieval

Positional Postings and Phrase Queries

Exercise 2.9: Shown below is a portion of a positional index: angels: 2: <36, 174, 252, 651>; 4: <12, 22, 102, 432>; 7: <17>; fools: 2: <1, 17, 74, 222>; 4: <8, 78, 108, 458>; 7: <3, 13, 23, 193>; fear: 2: <87, 704, 722, 901>; 4: <13, 43, 113, 433>; 7: <18, 328, 528>; in: 2: <3, 37, 76, 444, 851>; 4: <10, 20, 110, 470, 500>; 7: <5, 15, 25, 195>; rush: 2: <2, 66, 194, 321, 702>; 4: <9, 69, 149, 429, 569>; 7: <4, 14, 404>; to: 2: <47, 86, 234, 999>; 4: <14, 24, 774, 944>; 7: <199, 319, 599, 709>; tread: 2: <57, 94, 333>; 4: <15, 35, 155>; 7: <20, 320>; where: 2: <67, 124, 393, 1001>; 4: <11, 41, 101, 421, 431>; 7: <16, 36, 736>; Which docs if any match each of the following queries where expressions within quotes are phrase queries? a. fools rush in b. fools rush in AND angels fear to tread
58

Introduction to Information Retrieval

Positional Postings and Phrase Queries

Exercise 2.10: Consider the following fragment of a positional index: Gates : 1: <3>; 2: <6>; 3: <2, 17>; 4: <1>; IBM : 4: <3>; 7: <14>; Microsoft : 1: <1>; 2: <1, 21>; 3: <3>; 5: <16, 22, 51>; a. Describe the set of documents that satisfy the query Gates /2 Microsoft. b. Describe each set of values for k for which the query Gates /k Microsoft returns a different set of documents as the answer.

59

Introduction to Information Retrieval

Resources for todays lecture

IIR 2 MG 3.6, 4.3; MIR 7.2 Porters stemmer:
http://www.tartarus.org/~martin/PorterStemmer/

Skip Lists theory: Pugh (1990)

Multilevel skip lists give same O(log n) efficiency as trees H.E. Williams, J. Zobel, and D. Bahle. 2004. Fast Phrase Querying with Combined Indexes, ACM Transactions on Information Systems.
http://www.seg.rmit.edu.au/research/research.php?author=4

D. Bahle, H. Williams, and J. Zobel. Efficient phrase querying with an auxiliary index. SIGIR 2002, pp. 215-221.