|
|
|
We’ve just been talking about a collection
of online databases so far. Today, the
online database is truly the most common and delivery is often via the
Internet. CD-ROM databases were quite
popular in the roughly ten years before the advent of the World Wide Web, but
they proved difficult to work with – especially when it came to effective
networking. Tape loaded databases were
quite common before the Internet as well, but now they’re not used much. Sometimes disk drives are used to store
data, but again it is not common. The
online database accessed via the Internet is predominant. When I first started out in libraries, we
accessed host computers of online databases by dialing into them via phone
lines. Connection speeds were so slow
that I could read the data as it came across my screen. In library school, I used a dumb terminal
with what was known as an acoustical coupler.
We would type in our commands and then wait for a computer response to
printed onto the terminal paper – there was no screen. People used IBM Selectric typewriters as
their dumb terminals. My first supervisor
would talk of the days when she would use a keypunch machine to type search
commands onto Hollerith cards. The
cards would be mailed off to a university such as Georgia Tech and fed into
their computer. If there were any
typographical errors, you found out when the computing center would send you
your results via mail. This would take
weeks … even months. How things have
changed!
|
|
Databases can also be divided up by the
structure of their data. The
bibliographic database, such as ERIC or NTIS or a library catalog, are very
familiar to most people – especially to librarians and library school students.
|
|
The directory is also a common type of
database that one will experience.
They are just as tricky to search as other types of databases. Even a small directory of a few hundred
hotel guests or hospital patients can overwhelm those unprepared to use
them. At the recent ALA conference in
Chicago, I failed to reach a colleague because the query was too difficult
for the operator of the hotel. At
Carnegie Mellon, we routinely get calls for the public library because the
telephone operators have difficulty with the directory.
|
|
The fulltext database has now become very
common. Fulltext databases can be
difficult to search unless one can focus their search to bibliographic and
abstract data. They will often swamp
the database searcher with huge sets of search results. They can often be quite advantageous as
well. Once, I was trying to see if
anyone had ever found a way to put an electric charge on a bunch of tiny
pieces of glass. I was able to search
for the words glass near the word pellet or shard or bead or fragment that
also showed up near the phrase “electric charge” or “electrical charge.” I located a German patent that described a
way that a company did it in order to put a charge on particles in a sand
blaster – the idea was to have the glassy fragments of the sand blaster have
the dust from the blaster stick to the glass.
Remember in science class when you had to rub a glass rod with a sheet
of rubber? Well, the inventor had the
glass pellets pneumatically transported through a rubber-lined tube. The inner surface of the tube had helical
shaped ridges that caused the glass fragments to bounce all over the place –
the end result of the glass pieces bouncing off the rubber was a static
charge! That had to be one of my
happiest moments as a database searcher.
The engineer that I was working with was astounded when I found a
working method with just a couple of hours of planning and conducting a
literature search. We’ll be learning
how DIALOG lets us search through full text.
Then … take this to the bank … you’ll be looking for a good equivalent
search feature whenever you come across a fulltext database.
|
|
Citation databases – wow! Things have really begun to change in this
arena. Perhaps ten years ago, the only
citation databases were those in the Westlaw legal databases and the triad of
databases from the Institute for Scientific Information: the computer
versions of the Science Citation Index, the Social Sciences
Citation Index, and the Arts and Humanities Citation Index. The power of a citation index is that you
can start from one good article written perhaps … five years ago … and
quickly search for other more recent articles that listed the older article
in their list of references.
|
|
Financial databases are an interesting
monster! The key to searching them is
to gain an appreciation for the typical information found in a financial
record. Then, gain an appreciation for
how the data is structured – in DIALOG we would use their Bluesheets. Other database vendors also provide
searching guidance as well.
|
|
Numeric databases allow the user to do
things like search for materials with a certain set of properties … perhaps a
range of boiling points for a database of chemicals or metals with a
coefficient of thermal expansion below a desired value.
|
|
To help illustrate what the last three
databases are like, the URLs that I’ve provided will lead you to search aids
for databases of each type. The first
two are Bluesheets from DIALOG. The
third is the “Database Summary Sheet”
from the American Chemical Society’s STN databases. STN is a service that can be considered similar
to DIALOG. It has more of an emphasis
on scientific information while DIALOG seems to do an admirable job with all
types of databases (with the exception of numeric databases of scientific
information – note that my example wasn’t another Bluesheet).
|
