Prev | Current Page 781 | Next

A. F. Salam and Jason R. Stevens

"Semantic Web Technologies and E-Business: Toward the Integrated Virtual Organization and Business Process Automation"

Table 2 lists some of such products and gives an overview of
the strategies and characteristics of those products.
Impact.of.Spelling.Errors.on.Medical.Information.Seeking
Medical terms are usually very hard to spell. Typing incorrectly or spelling phonetically
can yield to inappropriate or irrelevant results, which causes a waste of time,
resources or, at times, can give a false alarm if the users are not aware of the error.
Blaster et al. (2004) found that spelling errors caused significant errors in patient
discharge letters. Through earlier usability testing on Australian medical portals
we found that the users did not want to come back to the sites that did not have
spelling error correction facilities to assist their search (Moon & Burstein, 2005).
In many instances the users could not find the information they were looking for
from such portals. We have concluded then that spelling correction facility is one
of the essential characteristics of an intelligent portal.
Table 1. Comparison of three spelling correction methods (Adopted from Kukich,
1992)
Techniques 521-Word Lexicon 1,142-Word Lexicon 1,872-Word Lexicon
Minimum Edit Distance 64% 62% 60%
Simple Correlation
Matrix (Hamming
Distance)
69% 68% 67%
Neural Network 75% 75% ?
Moon and Burste n
Copyright ?© 2007, Idea Group Inc.


Pages:
769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793