Eric Lease Morgan wrote:
>
> Computers may only ingest texts, but computer programs can statistically
> analyze texts, and through such analysis draw conclusions about said
> texts. This is what relevancy ranking is all about, and is the sort of
> thing that has made Google (and information retrieval techniques) so
> useful.
These same techniques can be applied to description of books and
> their "aboutness". I wrote a gentle introduction to this topic:
>
> http://tinyurl.com/n8k6bd
>
In that paper, you write
"
To many of us the phrase “relevancy ranked search results” is a mystery.
What does it mean to be “relevant”? How can anybody determine relevance
for me? Well, a better phrase might have been “statistically significant
search results”.
"
"relevancy ranked" it indeed a problematic phrase, not just a not so
good one. For those to whom ths concept is a mystery, the word
"relevancy" itself is probably less of a mystery but they will likely
understand or assume that some form of artificial intelligence is
at work here. For how else could a machine detect my intentions when
I tossed in those search words? Well, of course it can't. What's
relevant for me in that situation is subjective and must be a complete
mystery for the machine carrying out the search. Are all search engine
users aware of this?
So, "relevancy ranked" is a metaphor meaning something very different
from what the clueless searcher will likely understand. And that makes
it misleading. Something we better avoided because it is unhelpful and
potentially harmful.
In short: Avoid misleading metaphors.
B.Eversberg
Received on Mon Jul 13 2009 - 08:16:51 EDT