Searchers Toolkit, part 1

1
Searchers Toolkit, part 1

• Boolean logic, controlled vocabulary, field
  searching, and more

2
What this lecture will cover

• Boolean logic
• How parentheses affect the order of operation
• Controlled vocabulary
• Field searching
• Some terms in the searching lexicon

3
Basic Tool 1 Boolean Logic

• Boolean logic (after George Boole) refers to the
  logical (rather than arithmetical) operations
  on sets.
• The Boolean operators are
• AND
• OR
• NOT
• Note alternate syntax AND NOT

4
Boolean, cont

• Boolean logic is traditionally presented with
  drawings called Venn Diagrams

A
B
A
B
NOT
OR A OR B - either one, or both
AND A AND B - only items that have both
A
B
A NOT B - A only when B is not present
5
Order of Boolean Operations

• A statement using just one Boolean operator -
  ducks OR geese - is straightforward.
• But many searches require more complexity. What
  happens when you have more than one operator in a
  search statement? Generally
• NOT operations are performed first. Then AND
  operations are evaluated. Finally, OR. This is
  called the order of operation.

6
The Power of Parentheses

• Just knowing the order of operation doesnt make
  it very easy to write more complex search
  statements, however.
• Luckily, many systems also allow you group your
  ANDs, ORs, NOTs with parentheses ( ) - and
  either just make explicit the order of operation,
  or override it.
• Throwing parentheses into the mix can change the
  way the system interprets the search dramatically.

7
A 3-way Venn diagram
Here we have (A OR B) AND C What area would
represent (A OR B) NOT C ?
B
A
C
8
The Power of Parens

• Consider these statements ducks OR geese NOT
  migration (ducks OR geese) NOT
  migration(ducks AND geese) OR loonsducks AND
  (geese OR loons)(ducks OR geese) AND
  migration(ducks OR geese) NOT migration

9
Pause

• Human Venn Diagram activity

10
Basic Tool 2 Controlled Vocabulary

• The conundrum of online searching, and, in fact,
  most of librarianship We want to look for
  concepts, but we are forced to search for words.
  (Walker and Janes 1999)
• If a database offers a set of controlled
  vocabulary a.k.a. subject headings, thesaurus,
  descriptors, authority control - how does this
  help us?

11
Controlled Vocabulary

• Saves us from having to come up with every
  possible synonym for a term
• Gives us one way to look up, say, an author known
  by more than one name (think Mark Twain/Samuel
  Clemens)
• Can be used to disambiguate words that have
  several meanings (e.g. mercury - a planet? A car?
  A god? A metal?)

12
Controlled Vocabulary

• Theoretically makes our search more complete if
  we know that every time an article mentions
  geese, or goose, or duck, or loons, or any other
  water bird, it gets assigned the subject
  waterfowl - we just have to search on that
  term, and we should get everything.

13
Controlled Vocabulary

• In more formal terms, controlled vocabularies
• Facilitate the gathering of like items
• Help with comprehensiveness of results
• Help also with precision of results
• Help broaden understanding of a topic in an
  unfamiliar subject area (Walker and Janes 1999)
• Well get a better sense of what these statements
  mean when we see a database that uses controlled
  vocabulary in action.

14
Basic Tool 3 Field Searching

• Field Searching simply means the ability to
  restrict your search to a particular field, e.g.,
  look for a name in an author field.
• Field searching will focus your search, and
  usually make it more efficient.
• The database designers spent all that time
  deciding what fields to have - so definitely
  exploit this feature if it is available!
  (especially useful combined with controlled
  vocabulary) 1

15
Searching Lexicon 1

• False drops
• A false drop is a document that gets retrieved
  by your search terms - but the terms are not used
  in the sense you intended e.g., a search on
  employment or jobs or careers retrieving
  articles about Steve Jobs.
• False drops epitomize the problem of wanting to
  search Concepts but only being able to search
  Words!
• Controlled vocabulary and field searching help
  avoid the false drop problem

16
Searching Lexicon 2

• Stop words. (We mentioned these briefly in
  lecture 1)Stop words, a.k.a. noise words are
  those little words that most systems do not
  index, including but not limited to an, by,
  for, from, of, the, to, with
• Digging in the databases help files may reveal
  their list of stop words. Youll have occasions
  where this is very important!

17
Searching Lexicon 3

• Recall vs. Precision
• Recall and precision have to do with the number
  of results your search getsRecall refers to
  retrieving more results - spreading your net as
  wide as possible.Precision refers to focusing
  your search down, retrieving fewer - but more
  perfectly on-target - results.
• What might be the pros and cons of each?

18
End of toolkit, part 1

• So now you have 3 major tools in your search
  arsenal
• The concept of Boolean logic for combining terms
• The concept of controlled vocabulary
• The concept of field searching
• Lets look at EBSCO MasterFile Select, a
  multidisciplinary database, and see how these
  tools apply there.