Dialog can be daunting

I have used Dialog on-and-off since the mid 1980s, though most of my experience dates from 1995. Most of the time I used it to find research for scientists, engineers, and college instructors and administrators. It has always been a wonderful tool, even in its pure, text-based, telnet-accessed version.

It can be daunting, though, for two reasons:

• Cost
  • Only companies, universities, large research organizations, and information brokers (who are billing others for their costs) can typically afford the cost of retrieving and displaying records from Dialog. (Also, you can make searching errors that can inadvertently increase that cost dramatically.
  • So, I am going to repeat a number of times that you are going to have pursue a number of cost-containment strategies when you search, even in this class, because we have a limited allowance from Dialog for student searching each year.)
  • Remember that costs in the library that you work in can be offset by using other, non-Dialog-provided databases that are "free" (at least, you do not have to pay for them because you are accessing them through libraries that have licensed them for their patrons). Unfortunately, many of the databases in Dialog contain unique material not available or, at least, easily accessible elsewhere.
• Size
  • The number and diversity of databases in Dialog can sometimes make searching very challenging, particularly if the terms that you are searching on are common or have multiple meanings.
  • This is where Boolean, proximity, field searching, the use of tools to select target databases, and Bluesheets, the basic guides to using each individual database, come in.
  • Consequently, until all the content that is in Dialog's databases is available though other, particularly "free" aggregators, the database utility will continue to be a major player in the dissemination of electronic information around the world for those who can pay for it.

How to start out

1. Get a basic understanding of searching capabilities of Dialog. You are getting that in:
   • Chapters 6 through 10
   • My Dialog Web pages, which I have divided up into the various features and functions so that you can focus on individual aspects.
2. Try the step-by-step examples and tutorials for both DialogWeb and DialogClassic. This will give you a sense of how both Interfaces work.
   • Dialog's DialogWeb Online Introduction and DialogClassic Online Introduction (both .pdf files, 1.5MB to 2.5MB in size)
   • My DialogWeb Step-by-Step Example and DialogClassic Step-by-Step Example
3. Review the catalog and several Bluesheets to understand how they can be used to understand what databases may be relevant and how they can be searched.
4. Experiment, try some searches of your own, but always be aware of the costs and use the money-saving commands to constrain those costs.

Tips and Thoughts on Searching Dialog

The 80/20 rule

Never forget the 80%/20% rule. 80% of the time you will use less than 20% of the resources available to you. That includes search functions, databases, and ways of composing search statements/algorithms/strategies.

On the other hand, from time to time you will need to use a particular operator, a way of configuring your search statement, a database that normally you would not consider. That 20% or less of searching can often consume 80% of your time.

You don't have to be a subject specialist

You can be a successful searcher in almost any subject if you do just a few things:

• Acquire the basic subject knowledge that you need from your customer (do a thorough reference interview), from other print and online resources (such as the LCSH), and from experts.
• Learn as you search. Pick up synonyms and other cue words from the records that are displayed. Even if they are not totally relevant, they may contain terms, names, or other clues that will facilitate your composing an effective search strategy.
• Approach each search creatively with a willingness to experiment if what you are looking for does not come up.

Language becomes more important

In Dialog you have such a diversity of databases that selecting the terms appropriate to the kinds of records that you expect to turn up becomes more important. There are very general databases with lay, popular publications, to esoteric technical databases with arcane scientific or other professional language and, even, data formats (such as for statistical and chemical databases). There are also more chances of multiple, highly differing meanings for many more words. So, be very thoughtful about the language that you select.

And don't forget the Library of Congress Subject Headings/Authorities. Even though many of the standard terms in LCSH do not, shall we say, reflect ommon language, the majority of headings are reasonable. So, I highly recommend using the LOC's online Authorities Web pages (http://authorities.loc.gov/). Also use a good dictionary and thesaurus, such as the online Merriam-Webster ones at http://www.m-w.com.

Bluesheets

Bluesheets (http://library.dialog.com/bluesheets/) are descriptions of individual databases, the ways that they can be searched, and their cost. They are very useful tools, particularly when you are searching individual databases, but do not get hung up in them if what you are searching for is likely to be in a variety of databases from multiple publishers.

A particular problem, if you are doing field searching, is that databases vary in the availability, content, and designation of search fields. You may not have the time, particularly from a cost containment perspective, to deal with the idiosyncrasies of field-searching and other differences among databases.

Operators

I talk about these in other pages, but I like the old rule about going over a subject at least three times to reinforce the subject in one's mind.

"and" - In Dialog like other databases, "and" is still your most powerful search function.

Proximity operators (#W) (#N) - I have found that they are more useful in Dialog than in the popular database products. I am not absolutely certain why, particularly because there are many database products from the popular DB vendors in Dialog. But, be careful in the use of proximity operators, particularly the distance in number of words that you set, because you can eliminate excellent records relevant to your information need. It is a subtle juggling act, but, again, think about how we use language, how we relate words to each other in sentences.

Note: Phrase searching is also powerful in Dialog, but you use the "(W)" proximity operator by itself ( or just the parens () alone) between words to make a phrase.

Example: stephen(w)king and dark()tower.

Field operators (Prefix and Suffix indexes) - Also powerful in Dialog. Particularly useful are Suffixes: title (/ti), abstract (/ab), the standard descriptors (/de), and uncontrolled terms known as identifiers (/id). Field searching also increases the precision of your search, but, as noted above, remember that databases vary in what fields they contain and, even, how they may be designed.

Truncation is very important - You never know how something is going to be expressed in an article. For example, if somebody was writing about a method of measuring the heat loss of a container, they might use the word "measure" or "measuring" or "measured." Be suire to consider whether one, some, or all of your terms might benefit from truncation.

Be sure, though, that you don't truncate to a root that would result in too many docunents with non-relevant words being returned:

Example: the root hand? ( s hand? ) would return "hand" and "hands" but also "handle," "handy," "handling," "handsome," and more.

Solutions for above: s (hand or hands) or, more simply, ( s hand? ? ) just finds the records containing the words "hand" or "hands."

DialIndex (File 411)

One of your absolutely best tools for selecting databases. DialIndex enables a search of the union index (File 411) of all words in all Dialog databases for the frequency of occurrence of your search terms in those databases. Once that you have ranked those databases by the number of occurrences of your terms in those databases, you have a good idea of where to start, a good idea of which databases to limit your searches to.

On the other hand, every once in a while the most relevant record will occur in a database with a low frequency of the terms that you are searching for. For example, on a question about a twelfth century nun who composed liturgical music, the best answer came out of a local newspaper in a city where there was a concert of her music. It turned out that a professor at the local university just happened to be an expert on the nun-composer.

Which to use: DialogWeb or DialogClassic ?

Which interface to use? It all depends on what you are comfortable with. I like more control over my searches. This includes selecting individual databases and OneSearch categories to use with a DialIndex search and how I want to display the results of my search. Consequently, I tend to use DialogClassic. On the other hand, you can pretty much treat DialogWeb like DialogClassic because of what you can enter in the entry field at the bottom of the page.

DialogWeb gives you some guidance in the selections of databases to search because of its predefined categories. It also has predefined selections of output formats.