Information Retrieval

Adapted from Lectures by Berthier Ribeiro-Neto (Brazil), Prabhakar Raghavan (Yahoo and Stanford) and Christopher Manning (Stanford)

Prasad

L1IntroIR

Unstructured (text) vs. structured (database) data in 1996

160 140 120 100 80 60 40 20 0 Data volume Market Cap
L1IntroIR 2

Unstructured Structured

Prasad

Unstructured (text) vs. structured (database) data in 2006

160 140 120 100 80 60 40 20 0 Data volume Market Cap
L1IntroIR 3

Unstructured Structured

Prasad

Structured vs unstructured data

Structured data : information in tables
Employee Smith Chang Ivy Manager Jones Smith Smith Salary 50000 60000 50000

Typically allows numerical range and exact match (for text) queries, e.g., Salary < 60000 AND Manager = Smith.
Prasad L1IntroIR 4

Unstructured data
Typically refers to free text
qData which does not have clear, semantically overt, easy-for-a-computer structure

Allows
qKeyword-based queries including operators qMore sophisticated concept queries, e.g.,
find all web pages dealing with drug abuse
Prasad L1IntroIR 5

Semi-structured data
In fact almost no data is unstructured
qE.g., this slide has distinctly identified zones such as the Title and Bullets

Facilitates semi-structured search such as

qTitle contains data AND Bullets contain search
to say nothing of linguistic structure

Prasad

L1IntroIR

What is IR?
Representation
Keywords/Phrases, Structure/Fonts, Counts, etc

Organization and Storage

Inverted File Index, Compressed, etc Hardware Architecture and Memory Hierarchy

Access to information items

Interface : Spell-checker to tree-structured display Visualization : Labeled Clusters, Timelines, Spring graphs, etc.

Prasad

L1IntroIR

Ultimate Focus of IR
Satisfying user information need
qEmphasis is on retrieval of information (not data)

User information need : Examples

qPrinter reviews;Printer prices and availability
qWords in which all vowels appear qAnagram/Permutations of art qFlight numbers; UPS/FedEx/USPS Tracking code

Predicting which documents are relevant, and then linearly ranking them.
Prasad L1IntroIR 8

Information Need : Query, Relevancy

An information need is the topic about which the user desires to know more, and is differentiated from a query, which is what the user conveys to the computer in an attempt to communicate the information need. A document is relevant if it is one that the user perceives as containing information of value with respect to their personal information need.
Prasad L1IntroIR 9

DIKW Hierarchy
Data: Symbolic units
qE.g., Records of customer. qE.g., Bytes from sensors.

Information : Data with an interpretation (Who?, What?, When?, Where?).

qE.g., Records of current/new customer grouped by their ages. qE.g., Variation in temperature readings.
Prasad L1IntroIR 10

DIKW Hierarchy
Knowledge : Information organized with theoretical concepts or abstract ideas (How?)
qE.g., How many customers have cancelled the accounts in current fiscal year? qE.g., Analysis of temperature variation over the years and their causes.

Wisdom : Understanding of fundamental principles + Human Judgement

qE.g., What strategies can be employed to retain customers in the face of cheaper alternatives? qE.g., Global warming issues and the future of Earth.
Prasad L1IntroIR 11

DIKW hierarchy: Clark 2004

Context

Joining of wholes Formation of a whole Connection of parts

Wisdom
Future

Knowledge Information
Past

Novelty

Experience Understanding

Data
Gathering of parts
Prasad L1IntroIR

Researching Absorbing Doing Interacting Reflecting

12

You see things; and you say "Why?" But I dream things that never were; and I say "Why not?" George Bernard Shaw

Prasad

L1IntroIR

13

Information vs Data Retrieval

DATA: Unstructured : open to interpretation Usually incomplete or ambiguous (w.r.t. information need) Structured with well-defined semantics Well-defined semantics Exact match required - no or many results Foundations: Algebra/Logic Accounting
L1IntroIR 14

QUERY :

QUALITY OF Partial match allowed, RESULTS: relevance-based ranking

FOUNDATIONS: APPLICATION:

Probabilistic underpinnings Library

Prasad

User Task
Retrieval Database Browsing

qRetrieval
Purposeful HP Multifunction Printer Information

qBrowsing
Casual Big Bang, CBR, Element Genesis, Supernova, ... Hyperlink-based

qFiltering by Agents
Push Podcasts from B.B.C.s Naked Science
Prasad L1IntroIR 15

Logical View of Documents

Docs

Accents spacing

stopwords

Noun groups

stemming

Manual indexing

structure structure Full text Index terms

Abstraction (essentials)
qStructure, fonts, proximity, repetitions, etc
Prasad L1IntroIR 16

The Retrieval Process

User Interface

Text
4, 10

user need
Text Operations

Text
6, 7

logical view
Query user feedback Operations query

logical view Indexing

DB Manager Module 8

inverted file
Index

Searching
8

retrieved docs ranked docs

Prasad

Ranking
2 L1IntroIR

Text Database

17

IR Basics
Models and retrieval evaluation Query languages and operations
Improve inferring query context
(query expansion, relevance feedback)

Text operations
Improve gleaning of document semantics
(stemming keywords)

Efficient Access: Index and Search

qVisualization, Multimedia, Applications,
Prasad L1IntroIR 18

Clustering and classification

Given a set of docs, group them into clusters based on their content. Given a set of topics, plus a new doc D, decide which topic(s) D belongs to.

Prasad

L1IntroIR

19

The web and its challenges

Unusual and diverse documents Unusual and diverse users, queries, information needs Beyond terms, exploit ideas from social networks
qlink analysis, clickstreams, ...

How do search engines work? And how can we make them better?
Prasad L1IntroIR 20

More sophisticated semistructured search

Title is about Object Oriented Programming AND Author something like stro*rup
qwhere * is the wild-card operator

Issues:
qhow do you process about? qhow do you rank results?

The focus of XML search.

Prasad L1IntroIR 21

More sophisticated information retrieval

Cross-language information retrieval Question answering Summarization Text mining
Prasad L1IntroIR 22

Future Progress: Factors/Trends

Large, uncontrolled publishing media
qQuality issues

Cheap, fast and wide access

qEase of use (query formulation)

Variety and flexibility

qNavigational and Visualization aids qDirectory-based (Table of contents) vs Keywordsbased (Inverted File Index)
Index terms (automatic/human-created) vs Full-text

Privacy, Security, Copyright

Prasad L1IntroIR 23