Searching Periodical Databases

1
Searching Periodical Databases

• MIT 026B
• Winter 2002

2
Database Records

• Accession Number
• Title
• Author
• Journal
• Language
• Abstract
• Descriptors

Fields
Record
3
CONTROLLED VOCABULARY Authors Descriptors
4
NATURAL LANGUAGE Title Abstract
5
Truncation

• Natur

• Nature
• Natures
• Natural
• Naturalist
• Naturalistic
• Naturalism

6
Wildcards

• M?cDonald

• McDonald
• MacDonald

7
Positional Operators

• ADJ
• ADJ3
• WITH
• NEAR

8
Training cats and dogs as pets

• Search for articles about

9
Training cats and dogs as petsTraining DogsCa
ts Pets

• Search for articles about

10
List of Documents

• How to train a dog to kill people.
• The difference between dogs and cats.
• Training cats as pets.
• Training dogs as pets.
• Training your skunk to be a pet.
• Why dogs make good pets.
• How to train a cat to kill people.

11
List of Documents

• How to train a dog to kill people.
• The difference between dogs and cats.
• Training cats as pets.
• Training dogs as pets.
• Training your skunk to be a pet.
• Why dogs make good pets.
• How to train a cat to kill people.

12
List of Documents

• How to train a dog to kill people.
• The difference between dogs and cats.
• Training cats as pets.
• Training dogs as pets.
• Training your skunk to be a pet.
• Why dogs make good pets.
• How to train a cat to kill people.

13
List of Documents

• How to train a dog to kill people.
• The difference between dogs and cats.
• Training cats as pets.
• Training dogs as pets.
• Training your skunk to be a pet.
• Why dogs make good pets.
• How to train a cat to kill people.

14
Document/Term Matrix
15
Document Sets

• Train d1, d3, d4, d5, d7
• Dog d1, d2, d4, d6
• Cat d2, d3, d7
• Pet d3, d4, d5, d6

16
Dog AND Cat
d6
d3
d1
d2
d7
d4
dog AND cat d2
17
Dog OR Cat
d6
d3
d1
d2
d7
d4
dog OR cat d1, d2, d3, d4, d6, d7
18
Dog NOT Cat
d6
d3
d1
d2
d7
d4
dog NOT cat d1, d4, d6
19
Hints for Searching Electronic Databases

• Plan ahead
• Dont get discouraged
• Be prepared to repeat the search, using different
  strategies
• Use BOTH free-text AND controlled vocabulary terms

20
Recall vs. Precision
High
Building Blocks
Search Results
Recall
Citation Pearl Growing
Low
Low
High
Precision
21
Building Blocks Approach

• Identify the different concepts, or facets of
  the topic you wish to search
• Create a series of terms for each concept
• Combine the terms for each concept together into
  a single set, using the OR operator
• Combine the sets together, using AND, OR, or
  NOT.

22
Topic

• Ecological concerns in American poetry

23
Concept Map
24
Building Blocks Approach
Concept 3
Concept 1
Concept 2
Term 2a OR Term 2b OR Term 2c
Term 3a OR Term 3b OR Term 3c
Term 1a OR Term 1b OR Term 1c
ANDORNOT
ANSWER SET
25
Building Blocks Approach
Concept 3
Concept 1
Concept 2
ANDORNOT
26
Successive Fractions Approach
Concept 3
Concept 1
Concept 2
Term 2a OR Term 2b OR Term 2c
Term 3a OR Term 3b OR Term 3c
Term 1a OR Term 1b OR Term 1c
ANDORNOT
INTERIM SET
ANDORNOT
ANSWER SET
27
Fun with Your Thesaurus

• Novel (100 citations)
• Narrower terms
• American novel (25 citations)
• Bildungsroman novel (25 citations)
• Canadian novel (25 citations)
• Postmodern novel (25 citations)

28
Searching the term, Novel

• Novel 100

29
Exploding a Term

• Searching the citations for a term AS WELL AS the
  citations for all the NARROWER terms under it
• Novel 100
• American novel 25
• Bildungsroman 25
• Canadian novel 25
• Postmodern novel 25

200 Citations
30
Citation Pearl Growing

• Start with 1 or a few documents that are KNOWN to
  be relevant to your search.
• Investigate the descriptors and natural language
  terms in the documents and develop a list of
  concepts
• Use these concepts to find more terms, and
  collect more documents
• Repeat the process as necessary/desired.

31
Citation Pearl Growing
Title Author Abstract Descriptors
32
Citation Pearl Growing
Title Author Abstract Descriptors
Title Author Abstract Descriptors
33
Knowledge Growth
General Relativity
Quantum Theory
45
(quantum ADJ theory) AND (general ADJ
relativity) 45
34
Knowledge Discovery
Vibration Theory
Quantum Theory
(quantum ADJ theory) AND (vibration ADJ theory) 0
35
Knowledge Discovery

• The detection of patterns in large data stores
  which leads to the extraction of previously
  unknown, potentially useful knowledge.

36
Knowledge Discovery
Vibration Theory
Quantum Theory