Terminology Search

on the World-Wide Web

by Gabe Bokor

The World-Wide Web (WWW) is a huge potpourri of business

advertising, serious scholarship, pornography, political propaganda,
scam artistry, and many other things. No authority has ever
prepared a card system for the millions of sites or imposed any
standard of quality for posting information on the Web. Yet it is
easier to find even a single word on the WWW than in the best-
organized library. While experienced web surfers may have their
favorite specialized sites and on-line glossaries (Cathy Flick’s column
lists a number of them in each issue of the Translation Journal), the
usual point of departure for most research, including terminology
research for translation, is one of the many Web search engines.
Search engines are a magic tool for translators looking for an
elusive word, an abbreviation, or wishing to confirm a guess. It is,
however, important to employ the right search strategy in order to
take maximum advantage of the search engines’ tremendous power.

Search engines are Web sites equipped with special software

capable of periodically combing the Web and collecting each word in
a huge database. Most search engines also accept submissions from
Web site owners. Due to the different ways the different search
engines collect, process, and store this information, the results you
get from your search with different search engines may be widely
different. There are hundreds, if not thousands, of different search
engines on the Web, some of them regional, others highly
specialized. We’ll only make reference to the major general-purpose
search engines:

AltaVista
Yahoo!
Excite
Infoseek
Yahoo! is a “Web directory,” where information is stored under 14
major categories and a large number of subcategories. This engine
is ideal when you wish to obtain more information on the subject
and search using a common, well-known term. For example,
searching under the key word “UCLA” (University of California, Los
Angeles), Yahoo gave me 3 “category” hits and 863 “site” hits, while
AltaVista resulted in over 378,000 hits. This doesn’t mean that
Yahoo! is that much “smaller” or less complete than AltaVista, but
only that Yahoo! will give you the sites where UCLA is the key word,
while AltaVista will show you virtually all occurrences of the word (or
abbreviation) on the Web. You can make Yahoo search in the
“AltaVista mode” by clicking the “Go To Web Page Matches” button
next to your first search results, and you’ll get more additional
pages than you’ll ever want to leaf through.
Yahoo! can be used even if you don’t have a specific key word.
Just select the major category and go narrowing your field until you
reach the area you’re interested in. For example, if you wish to
know more about automobiles that use non-polluting fuels, you can
go to Science-Engineering, then Automobiles, and then Alternative
Fuel Vehicles, which will finally take you to Solar Vehicles, Electric
Vehicles, and Hybrid Vehicles.
AltaVista is the other extreme of the search engine spectrum. It
is ideal when you want to find the use of an obscure term or
abbreviation. It is my personal favorite, although you can
occasionally be overwhelmed by the number of hits you get, not all
of which relevant to your search. One nice feature of AltaVista is
that it can select the language of the page. When I was trying to
find the English equivalent of the German word Warenkunde, I knew
that there was no exact, generally used term for the same concept.
So I searched Warenkunde among the English-language pages, and
I actually found the personal page of a professor of Warenkunde,
whose title on his English-language page was given both in German
and in English, the latter as Professor of Commodity Science.
The same feature also came in handy when I tried to confirm the
use of the English terms “benchmark” and “benchmarking” in
Portuguese. Searching for benchmark or benchmark* in AltaVista’s
Advanced Mode among Portuguese sites, I got 1044 hits. It is
interesting to note that not all of the texts found used the term in
the computer context, but none (of those I checked) bothered to
provide a translation in the vernacular. I also confirmed that both
words are used in the masculine gender, which is not to be taken for
granted, considering that marca is feminine in Portuguese, and the
adopted foreign word often assumes the gender of its Portuguese
cognate or equivalent. For example, the word Internet is feminine
because rede (network) is feminine in Portuguese.
A note is in order about searching for phrases like Federal
Reserve Board as opposed to words. In most engines, if you just
type in the phrase, you get all the sites where any of the words
occurs. To restrict your search, you must either put your key phrase
in quotation marks or use the Advanced Option of the search engine,
which either allows you to use Boolean operators AND, OR, NEAR,
NOT, or gives you a menu selection to achieve the same result.
Obviously, searching for “Defense Department” (with the quotation
marks) will only bring up the exact phrase as typed. Typing Defense
Department (without the quotation marks) will give you pages
where either of the two words occurs in any context. defense and
department under advanced (Boolean) search will yield pages where
both words occur, for example, Defense Attorney on line 3 and
Justice Department on line 54. defense near department will yield
pages with Defense Department, Department of Defense, and
probably a few irrelevant sites where the two words happen to be
near each other. You must select your search strategy according to
the specific case at hand and the result you wish to achieve.
Excite is one search engine that claims to be context-sensitive;
i.e., if you enter car, it will also find sites where only the words
automobile or motor vehicle occur.
Most search engines are not case-sensitive. Punctuation marks,
including hyphens, and “stop words” such as and, or, is, for, etc. are
ignored unless they are part of a phrase between quotation marks.
Word fragments can be searched with some engines using the
asterisk (*) as a wildcard. For example, searching for annelat* (to
find occurrences of annelate, annelated or annelating, AltaVista
yielded 261 hits, Yahoo! 32 hits, and both Excite and Infoseek 0
hits.
Abbreviations are easily found on the Web. When a large
number of irrelevant hits occur or the expansion of the abbreviation
is not immediately found, the search can be narrowed by specifying
the language (with AltaVista) or using the known or suspected
portion of the abbreviation with Boolean operators. For example,
knowing that the first letter in the Portuguese abbreviation IPMF
stands for imposto (tax), searching AltaVista in the Advanced mode
for IPMF and imposto yielded 68 hits, many of which expanded the
abbreviation as Imposto Provisório sobre a Movimentação
Financeira.
The order in which the search engines list the hits is usually
random. Excite does it by what it considers order of relevance, even
giving percentage of relevance figures, but I’ve found this order to
have little to do with the actual relevance of the sites found.
Once you’re on a page that, according to the search engine,
contains the word or expression you’re looking for, it’s easy to find it
on the page. Just type Ctrl-F (Command-F on the Macintosh), and
you get a search dialog which allows you to find the string (word,
phrase, or fragment) on the page. You can also use the Find in Page
(Netscape) or Find (on this page) options under the Edit menu.
In doing terminology research using the search engines, you can
find the sites where a given term is used (and possibly its meaning
clarified); less frequently, you can find the translation of the term as
in bilingual or multilingual sites “Rosetta Stones” (such as those
listed in Cathy Flick’s column) or in cases like the one mentioned
above with Warenkunde. You can also confirm or disqualify your
guess of a term. For example, if you’re not sure whether the
German word Ventilhub in the automobile context is actually valve
stroke, you can search for the latter to find that the 77 occurrences
(in AltaVista) all refer to non-automotive context. Valve lift, on the
other hand, gives you 799 hits, most of them in the automotive
context. Using the search engines, you can find not only individual
words, but the type of lingo in which those words are used by those
who work in that particular field.
What I would not recommend is letting AltaVista’s Translate
machine translation utility do your translation. The sample below is the actual
translation, including unedited punctuation, of the first sentence of an Italian web
page, performed by AltaVista’s machine translation software Babelfish. A (fairly
literal) human translation is attached for comparison.
Italian Original AltaVista Translation Human Translation
Bisogna procedere coi piedi It must proceed with the lead Extreme caution must be used
di piombo quando si scrive (e feet when law) of "electronic " when writing (and reading)
si legge) di biblioteche libraries, " " virtual " digitalises " about "electronic," "digital," or
"elettroniche", "digitali" o is written (and or, terms that are "virtual" libraries, these terms
"virtuali", termini che si wasted, of these times, also in being liberally used today even
sprecano, di questi tempi, the reviews and the television in the most popular magazines
anche nelle riviste e nei programs more divulgativi, and TV programs, while it is not
programmi televisivi più without that is always clearly to always clear to the reader (and
divulgativi, senza che sia who law (and sometimes to who sometimes not even to the
sempre chiaro a chi legge (e does not write) to what is author) what is actually being
talvolta nemmeno a chi effectively reporting to us. discussed.
scrive) a cosa ci si stia
effettivamente riferendo.
.
If the major search engines don’t give you the information you
need, you can use them to access other, regional or specialized,
search engines such as the Brazilian Cadê, the European EuroSeek,
the Japanese Info Navigator, the Russian search engines listed in
Cathy Flicks column in this issue of the TJ, or the Medical Search
Engine. There is at least one for each need, taste, language, or
specialty.

© Copyright 1998 Translation Journal

Send your comments to the Webmaster
URL: http://accurapid.com/journal/07search.htm
Last Updated: 09/27/2000 20:36:37