P.S.: I’m starting to get good at adding Roman numerals.

I am planning to change it soon. Maybe go for mayan numerals?

http://en.wikipedia.org/wiki/Maya_numerals

or Egyptian…

http://en.wikipedia.org/wiki/Egyptian_numerals

I’ve been using it for about six months. It’s designed to be used in a very interactive and manual way. You’re supposed to iteratively refine your query and also to go through the list of hits and remove the check marks from the ones that don’t belong. It’s not fully automatic, but at least the user interface makes it relatively easy to refine the results.

P.S.: I’m starting to get good at adding Roman numerals.

What the Web taught us is that recall is far more important than precision.

The most common problem I have is that I missed a reference. This can cost you months of research time. Google Scholar may be biased (aren’t all libraries biased), but it has excellent coverage.

At least, it indexes my papers. I bet that if you looked in CISTI’s database, you’d find, at best, half of my papers. Yes. It matters: if you are one of my competitors, and you missed half of what I have done, you are basically flying blind.

I have this feeling that librarians of the future will basically be bibliographic data-analysts. Just today I found that one of our databases had indexed many articles several times (i.e. separate entries and differnt journal volumes etc. for the same article), no doubt because of some software error that arose in in the creation of the DB.