[Web4lib] RE: Another Google question

[Mike Taylor]

> Date: Fri, 15 Jul 2005 11:04:42 -0700
> From: Karen Coyle <kcoyle at kcoyle.net>
>
>> The big, big win you get from Google is that its top hit (or
>> second, or third) is nearly always the one you want.
>
> What Google is very good at is the retrieval of pages based on
> proper names (the name of a company, an organization, or a product)
> where there is a single obvious answer.

This is a very apposite observation.  Yes, you're right; now that I
think about it, that is indeed where Google is strongest, and I'd
never really spotted it before.  Which I guess just goes to show that
_most_ of my searching is of this kind (and I further imagine that the
same is true of people users).

> For other types of searches, Google doesn't work so well. There's no
> "conceptual" searching. Topics like "childhood development" or
> "legal theory" come out very poorly. For names of people, you tend
> to get pages that have lists of the graduating class of blah blah
> high school, because they have every possible forename and surname
> combination.

All true.  I'd not want people to think I am a Google cheerleader --
my point is that it's perfect, or even that it's "good enough", but
that if everyone except information professionals _think_ it's good
enough (as they seem to do) then there's very little point in our
complaining, in our little info-ghetto, about weaknesses no-one else
cares about.

That said, here are a couple of research topics I have absolutely
_nowhere_ with using Google.

- I have been trying very hard in the last week to find a formula that
  shows what angle a cylindrical rod can be bent through by a force,
  given the force itself, the rod's length and radius, and its Young's
  modulus.  You'd think this would be trivial to find, but I've had no
  success at all.

- I want to find the compressive strength of hyaline cartilage (the
  kind of cartilage that caps your long-bones).  Again, no joy.

- A while back I was trying to find a Thunderbird 2 module for my
  (then) five-year-old.  I knew that the kind I was looking for is
  about 15 inches (38 cm) long, but that information is no use in a
  Google search.

It would be great for a search engine to find a way to handle this
kind of query.  I don't see it happening any time soon, though
... especially if it's going to be dependent on the Semantic Web :-|

> (Note: you and I do well on google searches because not only do we
> have our own web pages, we have our own domains. I suspect that
> makes a big difference.

Yeah; but, still, for a long time I was listed behind the Mike Taylor
who was running to represent North Carolina's 8th congressional
district back in 2000, even for some time after his campaign had
finished and the site had stopped being updated.

(Maybe I would be less aware of Google's strengths if I wasn't it's
top Mike Taylor :-)

> One of the things that I think we can conclude about Google, and
> that it should be possible to study, is that people tend to approach
> search engines with a verison of the library world's "known item"
> search.

True.

 _/|_	 ___________________________________________________________________
/o ) \/  Mike Taylor  <mike at miketaylor.org.uk>  http://www.miketaylor.org.uk
)_v__/\  "No no, no no no no, no no no no, no no; there's no limits" --
	 some bunch of talentless nobodies.

--
Listen to free demos of soundtrack music for film, TV and radio
	http://www.pipedreaming.org.uk/soundtrack/