Monday, August 8, 2011

Metadata Monday: Stemming

Are you familiar with stemming? In linguistics, stemming is essentially reducing a word to its root. For example, the word 'fish' is the stem of fishing, fisher, fisherman and others. "Discuss" is the root of discussion and discussing.

Stemming algorithms have long been used in search and SEO to increase the results associated with any given term. Keyword stemming involves the strategic use of synonyms and alternative keywords. Here is a great list of tips for using this concept in SEO.

Until last week, it did not occur to me to employ this strategy to my advantage in planning the controlled vocabulary behind the DAM system I work with. Our keyword field specifically has been littered with all forms of nouns and verbs increasing the size of the list.

I did some research on the configuration of our DAM and it turns out that one simple line in the search file, Stemming = TRUE (or FALSE), was all that was needed to ensure that if we entered the term baby, the search results would also include babies. If we entered farm, we would also get farms and farmers and farming. It is like magic, our keyword list is more concise and precise as a result and consistency improves as well.

 

No comments:

Post a Comment