Product: EndNote Plus 2.1.3 for the Mac
Author: Torsten Heycke, Symbolic Systems Resources Group, Stanford University School of Medicine
Discipline: Medical Informatics, information dissemination (glorified technical writer)
In the
year 0 AW (After Web) the Knowledge Systems Lab and the Section on Medical
Informatics at Stanford University began disseminating technical publication
references via the World-Wide Web using EndNote as a means of generating
the HTML. Since that initial effort we have greatly expanded our offerings:
we not only include citation information but also the abstract and,
leveraging off the structured nature of the EndNote database, we offer
a structured field search capability.
In early
1994 the web was in relative infancy. Text-to-HTML converters were being
created rapidly, but had not been refined enough to reliably convert
data. We had already created a substantial EndNote library of our technical
publications (550+) and it seemed natural and expedient to experiment
with the EndNote styles facility to see if we could quickly produce
HTML. This proved to be relatively easy. We decided to further transcend
the conventional use of EndNote by entering full abstracts along with
citation information.
At about
the same time, we became aware of the structured field querying capabilities
of the free-WAIS-sf UNIX program and its Web interface (or gateway)
called SF-Gate. Simply stated, this meant that if we could mark-up our
technical publications in such a way that the all the fields could be
distinguished from one another, we could create a Web database search
that allowed one to search by one or more fields. Once again, because
of the structured nature of EndNote and the ease of editing the styles,
we were able to set up our database to be used by the WAIS-sf program.
(Basically, we used a unique comment tag before and after each field.)
We could
now search not the only full text in our database, but also by the author
field, the title field, the publication type, keywords, notes, etc.
This is not unlike the facility provided with EndNote itself, except
that you get the full power of the WAIS search engine including result
ranking, boolean searching, etc.
There are,
of course, many technical details that I have not described. The conversion
from Macintosh text files to UNIX text files offers several interesting
challenges. The insertion of HTML tags wreaks havoc on certain EndNote
special characters, e.g., the |. The EndNote HTML files need to be parsed
by a PERL script on the UNIX server to create indices by author, report
number, year published, etc. We also implemented an automated cross-checking
script to detect the availability of a PDF (Portable Document Format)
full-paper files and amend the existing HTML file to reflect this availability.
To see
our technical reports, visit Section on Medical Informatics Technical
Reports web page at http://www-camis.Stanford.EDU/pubs or the Knowledge Systems Lab Informatics Technical Reports web page
at http://www-ksl.Stanford.EDU/publications. You can view the source of any abstract; what you see is what EndNote generated for us.
|
