I am looking for a partner in commercializing an efficient information
retrirval engine for SGML/XML data we have developed recently.
The engine employs BUS (Bottom Up Scheme) technique that I devised, by
which
indexing overhead can be minimized. The basic idea of BUS is that
indexing is
performed only at the lowest levels of the document structure, while
retrieval
at a higher level can be done at run time by gathering index information
made
at the lowest levels. From an experiment, the index overhead amounts to
30 percent of the source documents tagged in SGML and the retrieval
time is fast as well.
BUS allows the element search and retrieval at any level in the document
structure including:
(1) nested structural search (searching elements that are ancestors of
another elements)
(2) content search
(3) element traversal at an arbitrary level
(4) retrieval of the elements in documents corresponding to a specific
element in a DTD
As for the BUS technique, we demonstrated as well as published a
technical paper
in the Digital Libraries '98. The response of the demonstration was so
good.
If you have an interest, you can test it with a JDK 1.1.5 enabled
browser in
http://savage.comeng.chungnam.ac.kr/~sgml
In this site, you can download a technical paper as well as test the
demo system.
Any inquiry about the system or cooperation is welcome to:
E-mail : shin@comeng.chungnam.ac.kr
FAX: +82-42-822-4997
Postal mail:
Dongwook Shin, Professor
Department of Computer Engineering,
Chungnam National University
220 Kung-Dong, Yusong-Gu 305-764
Republic of Korea
Thanks for your reading
Dongwook Shin
-- Dongwook Shin Department of Computer Engineering, Chungnam National University, Korea E-mail: shin@comeng.chungnam.ac.kr FAX: +82-42-822-4997 URL: http://savage.comeng.chungnam.ac.kr/~shin/index.html