[Sigia-l] Google Scholar

[Karl Fast]

> What value does it add, looking at Google does not own any database
> of Scholarly content? 


Google Scholar does to the academic literature what Google News does
to current events information.

It adds value in a number of ways.

The scientific literature has a specific and explicit structure that
could be algorithmically identified and extracted. The big items
would be names of authors, institutions, journals, conferences, and
paper titles.

There is also the reference list which you can cross-reference
against the rest of the database.

By subsetting this material from the master index, you can develop
some unique features and ranking algorithms that exploit this
structure. Google provides a master interface to the "universe of
knowledge" but sometimes you want to start by filtering out a large
portion of that.

You can also add search features relevant to this subset without
polluting the master set. For example, you can do a name: search in
Google scholar to find papers by a particular author. They could
develop a rich search syntax relevant to scientific literature
without it confusing the normal Google search syntax.

This is not a new idea.

Citeseer already crawls the web for scientific literature and builds
a searchable index. Very useful. I use it regularly.

   http://citeseer.ist.psu.edu/ 

Google Scholar seems to be a competitor to this. Some diffs

  - It doesn't offer the same feature set as Citeseer (yet).

  - I don't know how the sizes compare. Citeseer has been around
    longer, but Google has more resources

  - Google Scholar will be more reliable and accessible. Citeseer
    is occassionally unavailable because too many people are using it

  - Google has a branding edge, as mentioned


  

--karl
http://www.livingskies.com/