Using a Thesaurus to Create a Vocabulary - a Monster of a Problem

Generating synonyms is a big part of creating a good vocabulary, but if you think using a thesaurus - or building one into your keywording software - is a quick and thorough answer, think again.

The basic problem is that many words which are proposed as synonyms are pointless, misleading or undesirable, if left unedited

Consider, for instance a simple word such as bicycle.  A major thesaurus generated these synonyms for that word:

bike, cycle, tandem, two-wheeler, velocipede, wheels.

Just loooking through this string there are obvious difficulties.  A "tandem" is a bicycle built for two, but someone searching for an image of a tandem would not to be pleased to see many images of bicycles with only one person aboard.  The word "wheels" seems an odd choice as it is not in itself a strict synonym.  "Velocipede" is okay, although there wouldn't be many people searching with that word.

Much bigger headaches start with more conceptual terms such as "romantic".  Here are the thesaurus-generated synonyms we got for that:

amative, amatory, aphrodisiac, ardent, attached, boy crazy, doting, enamored, erotic, fond, girl crazy, have a crush on, horny, hot, hot and heavy,impassioned, in love, infatuated, lovesick, lovey dovey, lustful, passionate, romantic, sexy, sweet for, sweet on, tender, turned on

Imagine the image is of a couple in their 70s having a tender, loving kiss on a beach.  Now that's romantic, but do you think that "girl crazy", "aphrodisiac" or "sexy" really capture the meaning and essence of that shot?  The problem is that words such as "romantic" have many different shades of meaning, some of which can be conflicting.  If this issue is not addressed, the value of keywording is severely weakened.

Sad to say, as we look at keywording around various sites we find a lot of these curious and problematic "synonyms" that have not been edited when the vocabulary was first created, or at the time keywords were entered.  You have to suspect that this is due to either laziness or a lack of understanding of the English language, or both.

If developing your own vocabulary, we suggest you edit all entries for the precise shade of meaning before adding them to the vocabulary.

Alternatively you can rely on excellent keywording staff to edit down these words during entry.  However, this does require a high standard of English knowledge.

For companies which just automatically load an electronic thesaurus into their software and take whatever words are generated, all we can say is that you should examine whether you aren't short-changing your customers.