Does Google recognize character entity equivalences?

If your content contains a word that might be spelled with an accented
character -- for example, an 'e acute', such as "cafe" or "resume" --
you can choose to display it with the accent over the 'e' by using an
HTML character entity. In this case, the e acute entity would be either
é or é which are supposed to be equivalent.

In other words, if I wanted to display the word "resume" (meaning an
employment-history outline, rather than the verb meaning 'to
re-commence an action' which doesn't get an accent), I could code it
as:

résumé (more French-like spelling with two accented e's)
or
resumé (more common English spelling with one accented e)
or
resume (most common spelling online)

Same issue with words like cafe, blase, etc.

I was trying to figure out whether Google, or any other search engine,
uses a table of correspondences for these character entities, so that
(for example) for search purposes é = e. As far as I can tell
they don't: when I search on the same word or phrase but using
different forms of those letters, I get different numbers of hits.
However, I can't easily tell whether some of those entity forms are
just treated as subsets of others in their equivalence tables. In other
words, do the 2.6 million hits for "cafe + coffee" include the 1,510
hits for "café + coffee", or are they entirely separate sets?

What do people do if one spelling seems more common in Google's search
results, but also looks slightly illiterate? Do you go for the more
common spelling or the more literate spelling?

Thanks much!

--
Lawrence San
Cartoon Stories for Thoughtful People:
<http://www.sanstudio.com>
email: san (AT) sanstudio (DOT) com