Dear All,
The LAR paper by Rachel Hammond raises a number of interesting points, e.g.
cross-database searching, natural-language to keyword mapping etc.
There are however insufficient details of the 'SignPost' method to judge its
quality and compare it to existing search / mapping algorithms. UMLS
(mentioned by Ms Hammond), SnowMed, and 'hyperintelligent' search engines
such as the one planned for the NeLH, which I believe will have in-built
quality filters and go far beyond material contained in one-dimensional
electronic databases, look much more promising.
Rachel Hammond is to be congratulated for being brave enough to enter the
lion's den and suggest solutions for reliable cross-database searching, but
the article does not describe a software tool 'that might make every medical
librarian's dream come true'.
Medicine is too complex and fuzzy a subject to (?ever) yield its knowledge
base with one-step/stop searching.
Reinhard Wentz
Reinhard Wentz, Dipl. Bibl.
Imperial College School of Medicine
Medical Library
Chelsea & Westminster Hospital
369, Fulham Road
London SW10 9NH tel.: 020 8746 8109
e-mail: [log in to unmask] <mailto:[log in to unmask]>
|