Scope

Okay, you know these.  The only difference is that in Dialog you need to use "but not" for "not."

"As in most searching, "and" continues to be your most important tool. But in Dialog or other collections of simultaneously searched databases:

• "or" becomes more important because, unless you are looking for a specific kind (professional, scientific, lay) of language, you need to account for a greater range of ways of expressing the same concept.
• "not" (or "but not" or "and not") becomes more important because in large collections of records you have irrelevant content to exclude. This is particularly so in search engines with billions of files. Again, though, a warning that misuse or over-enthusiastic use of "not" can eliminate relevant records.

 

Boolean	Format	Examples, Notes
and	and	cats and dogs
or	or	cats or dogs
not	but not	cats but not dogs

Scope

• In every language, every dialect, every social subgroup, every profession, there are common:
  • Ways of constructing sentences.
    • For example, common sentence constructions in English include:
      • noun-verb
      • article-noun-verb-adverb
      • noun-verb-adjective-direct object
  • Words and phrases and abbreviations
    • Those common to a culture, community, or social group
      • bloody (Anglicism)
      • know what I mean (hip-hop, rap)
      • heavy heart
      • geek speak
    • Those common to a job, profession, or avocation.
      • diabetes mellitus (medicine)
      • charm (a characteristic of the sub-atomic particles called "quarks")
      • home run (baseball)
      • Society for Creative Anachronism
• For any subject, there is a probability that certain words or phrases will occur within :
  • Close proximity to each other
    • Example:
      • If you are looking for articles on the breeding of horses in Kentucky, you might find:
        • Kentucky has numerous horse breeding facilities.
        • Kentucky has numerous facilities for the breeding of Arabian horses.
        • Trotters are bred on small farms throughout the remote counties of Kentucky.
          • Note that there is not use of the word "horse.
  • Within the same sentence, paragraph, or page
    • Example:
      • An article on Macintoshes will probably have words like "Mac" and "performance" in the same paragraph as "G5" (the microprocessor) and "processor" and "2GHz" (the clock speed of one G5 microprocessor.
      • In an article on research on the thermal analysis of houses for leaks may have the words "heat," "wall," "windows," in relatively close proximity to each other. How close is the problem.

Consequently:

• To search effectively, you need to:
  • Have a general idea the way language is used:
    • In any social group.
    • In different professions.
    • In different fields or avocations
    • In various types of publications, including;
      • General magazines and newspapers
      • Professional journals
      • Subject-specific magazines
      • Academic journals
  • Know or find the words (and synonyms) appropriate to the subject you are searching for.
  • Know the available proximity operators for a database and how to use them.

Proximity operators

• Proximity operators are letters and/or symbols that enable the searcher to specify how close to each other she/he expects selected terms to occur.

List of operators

Function/Command	Format	Examples, Notes
adjacent in order entered	(w)	computer(w)systems Finds phrases such as "computer systems.
near in order entered	(#w)	computer(6w)systems # can be any suitable number
adjacent in either order	(n)	search(n)example Finds "search example" and "example search"
near in either order	(#n)	dog(9n)cat Finds "cat .... dog" and "dog .... cat"
within same field	(s)	teacher(s)student Both terms must be in the same field in either order
within same descriptor pair (heading-subheading)	(l)	solar(l)energy i.e., heading(l)subheading

Introduction

There are various estimates of several hundred thousands to as many as 1.8 million different words used in the ecumenical and welcoming embrace of the English language. Of course, this includes numerous foreign words that have been adopted (pho, tatami, bangra, chocolate, avocado, llama, and endless more) and a seemingly limitless number of scientific/medical/technical terms (normotriglyceridemic abetalipoproteinemia, Van der Waals bond, icosadodecohedron, serial storage Architecture).

So, how do we navigate through such a rich vocabulary when we are searching. I provided a clue in the last set of examples: phrases. Searching for an ordered sequence of words is one of the most powerful searches that you can do because of the increased precision you achieve when your are looking for a logical and expected phrase appropriate to your search. There are many kinds of phrases that are useful for searching. These include names of people or places, quotes, titles, scientific names, and much, much more.

But, you must be aware that in Dialog:

• Some fields are "phrase-indexed," which means that the content of the particular fields that are indexed must be searched for as a phrase. You are not searching for individual words.
• Other fields are "word-indexed," meaning all individual words and their locations in each field have been indexed. Words in these indexes can also be searched for as phrases, but you must use proximity operators to do so.
• You must look at the Bluesheets for individual databases show how the individual fields are indexed, either by word or by phrase.

Doing Phrase Searching in Dialog

In Phrase-Indexed Fields:

• In phrase-indexed fields, you merely need to enter the content as it should be displayed in the field
• Example: S au=swift, jonathan

In fields that are Word-Indexed

• In word-indexed fields, you usually use the (w) operator or just the parens () alone to create a phrase that you want to search for.
• Example: ss jonathan()swift and english(w)satire

Introduction

• A travel writer rode on an inflatable boat down the Colorado River. She describes the water in a certain section of the river:
  • "The water flows smoothly in this broad swath of the river."
  • "The water is flowing smoothly, lulling us as we drift."
  • "The water flowed gracefully, lacking any impetus to rush."
  • "We watch the water flow idly by."
• You never know for sure how someone will use a word, much less write a single sentence, and, even more, how they will use language throughout an entire document. English is a rich language, offering us multiple ways of expressing ideas, even in variations of one root word. Consequently, the use of symbols ("wild cards" & "Truncation operators") that stand for one or more letters enable us to search for variations of a root word or multiple spellings of a word can be a valuable tool in online research.
• In particular truncating a word is a powerful tool in Dialog searching which increases your recall because it addresses the possible ways a writer may employ a single word that may be vital to your search need.

Operator

• Dialog provides one wild card/truncation operator, the question mark (?).
• The question mark, though, can be used in a variety of ways to specify one or more unknown letters:
• Within a word (wild card)
• At the end of a word or root word (truncation).
  • Warning: Do not truncate words too far as you may produce a far too broad set of largely irrelevant results. For example, if you want to find out about how a company is computerizing its packaging operation:
    • Good: s computeriz? and packaging - This finds computerize, computerizing, computerized, computerization.
    • Bad: s computer? and packaging - This finds the computerize variants, but it also finds computer, computers, computerdom, computerless, etc. Just the words "computer" and "computers" would produce a huge set of irrelevant results.

Formats 

Function/Command	Format	Examples, Notes
truncation: any # of characters	?	flow? finds: flow, flows, flowing, flower, flowed, flowmeter, etc.
truncation: exact # of characters	??, ???, etc	catalog?? finds: catalogue, cataloged, catalog, catalogs, etc.
truncation: 1 additional character	? ?	cat? ? finds: cat, cats, Cato
embedded wild card: exact # of characters	??, ???, etc	wom?n finds: women, woman, womyn princip??s finds: principals, principles

Scope

• Dialog's search-and-retrieval software has rules for the specific order of execution of operators.
• The enables the searcher to control the exact manner in which the search is conducted.
• The most important things to remember is that anything inside of parentheses are executed first.

  Example: ss engineering and (bridge? ? or span? ?) but not united()states

  • In this example, the words "bridge" and "span" (and their plurals, though "bridge? ?" also finds "bridget" and "bridged") are searched for individually first.
  • Then each of the words is individually combined with the word "engineering."
  • Finally, all records that contain the phrase "united states" are excluded.

Order of Execution

1. Anything in parentheses ( ) (parenthetical phrases)  
2. (w), (n), (s), (l), or (f) (proximity operators)
3. not 
4. and 
5. or

Introduction

• "Nesting" with parentheses ( ) can improve a search in several ways, including controlling the order of execution of the terms that are being searched on, establishing specific relationships between terms, and the provision of choices when you are unsure of specific terms that might be used in a record.
• In my opinion, searchers do not use this function enough.

Usage

• "Nesting" enables the
  • Searching for terms in a certain order
  • Combination of searches into more complex search strategies.
  • Searching for:
    • Alternate spellings of a word:
      • Example: (catalog or catalogue)
    • Alternate words with essential the same meaning.
      • Example: (car or automobile)
• Nesting uses Parentheses ( ) to control the order of execution
• Specifically, words surrounded by the parens will be search on first.
• Examples:
  • (horses or cows) and farmer
  • horses or cows is searched on first, then each word is combined with the word farmer.
  • In other words, the search is equal to:

    horses and farmers and cows and farmers

• Nesting is also useful when you do not know the exact terms that might be used in a record.
• Example:
  • (Shetlands or horses) and farmer
• In some databases there can be multiple levels of nesting
  • Example:
• ((shetlands or horses) (3n) farmer) and (ireland or scotland)

Scope

• You sometimes may want to limit your search to a range of dates for several reasons, including:
  • You will better target your search if you are sure of a reasonable range of time during which a record relevant to your information need is likely to have been published.
  • You may want up-to-date information, such as medical info.
  • You may want information from a specific time period, such as articles contemporary to the the attack on the World Trade Center.
• The most important aspect of date limitation is that it increases the precision of your searches.

Ways of limiting your search

• pd - publication date
• py - publication year
• current# - current year plus # of years before the current year

Publication Date (pd=)

• This limits your search results to a specified date or range of dates.
• Format:
  • Single date:
    • s pd=yymmdd
  • Range of dates:
    • s pd=yymmdd:yymmdd
• Examples:
  • ss impeach? and president and pd=981200
  • ss blackbird and crash??? and pd=951007
  • ss s2 and hurricane? and pd=920800:921100

Publication Year (py= or /py)

• This limits your search results to a specified year or range of years.
• Format:
  • Single date:
    • ss <set or terms> and py=yyyy
    • s <set or terms>/yyyy (py is not required when used as a suffix)
  • Range of dates:
    • s <set or terms> and py=yyyy:yyyy
    • ss <set or terms>/yyyy:yyyy (py is not required when used as a suffix)
• Examples:
  • ss blackbird(3w)crash? and py=n number of back years back year.
  • Use it with the begin command when you search single or multiple files, OneSearch categories, or a combination of OneSearch categories and files.  
  • Formats
    • current1 results include the current year plus 1 back years
    • current2 results include the current year plus 2 back years
    • current3 results include the current year plus 3 back years
    • current4 results include the current year plus 4 back years
    • current5 results include the current year plus 5 back years
  • show current command displays the active current command.
  • To cancel current:
  • Enter the active current command followed by the word cancel, for example, current3 cancel.

Introduction

LimitAll (lall) is useful for reducing the number of records that are processed and retrieved for all subsequent s or ss searches until you cancel the Limit All command.

It is useful for both increasing precision and containing costs, the latter mostly by limiting the time involved in the execution of the search.

Usage

• The limit all command (lall) restricts subsequent select or select step searches to one or more limit suffix codes.
  • Example:
    • lall/ger - This limits all subsequent searches to records that are in German.
• Multiple one suffix codes can be entered separated by commas.
  • Example:
    • lall/ti,ger,de - This limits all subsequent searches to the title and descriptor fields and records that are in German.
• A limit all is cancelled when you use any of the following:
  • begin (b) command when changing a database
  • logoff
  • limit all cancel commands
    • lall cancel
    • lall-
    • lall/all
• NOTE: limit all cannot be used with OneSearch.

Example

Function/Command	Format	Examples, Notes
limit all	lall/ suffix codes	lall/ger, de

Scope

• Stop words are words that are not searched on at all by a database, though it is important to note their possible or probable existence in a phrase, as their position(s) can influence how one should construct a search. See the "The Old Man and the Sea" example below.
• There are a number of common stop words that are not searched on in Dialog, including the Boolean operators and, or, and not and the other words in the table below.
• Note that punctuation is also not searched on.

Stop Words

an	and	by	for
from	of	or	the
to	with

Problem: How do you search on a phrase that includes stop words?

• Use the (w) or () operator.
• Examples:
  • A search for "The Old Man and the Sea" could be done as follows:
    • s old(w)man(2w)sea

    A search for "U.S.A" could be done as follows:

    • s u()s()a

Scope

• keep enables you to combine selected records from different sets into a single set.
• This is useful when two or more individual searches produced certain records relevant to your information need.
• The new keep set is numbered S0 (set zero).
• You can subsequently use the keep set just like any other set , including selecting ( s ), typing ( t ), displaying ( d ), etc.

Usage

• keep can be used with
  • Set numbers
  • Set numbers and specific record numbers
  • Dialog accession numbers
• If you do not specify a set number, the keep command defaults to:
  1. The last set typed (t) or displayed (d).
  2. The last set (sn) number.

Usage in OneSearch

• In OneSearch, you can use keep with records or sets from multiple files.
• Formats:
  • keep <set number> from <file number>
  • keep <set number>/<record numbers> from <file number>
  • keep <set number>/<record numbers> and words from <file number>

Cancelling keep

• keep can be canceled several ways. Remember that cancelling keep will also remove the keep set.
  • keep cancel
  • k-
  • Entering logoff or begin ( b ) commands

Example

Function/Command	Format	Examples, Notes
keep	k Sn/n,n-n Sn/n,n	k s3/2,9,23-15;s6/3,4 from 275 k s3/22,29 from 275

Scope

• You can speed your searching by stacking commands.
• Stacking consists of entering two or more commands in one line separated by semicolons (;).

Some Commands that can be stacked:

begin	expand	execute	exs	find
keep	limit	rd	repeat	save
save temp	select	select steps

Example

b 85,275;ss (internet(w)research)(8n)library