Ontology-based search has potential
in that the listed hits are relevant to specific domain, thereby reducing the retrieval
of irrelevant word lists.
Conclusion
Natural language processing is complicated, and English is a challenging language
when it comes to spelling. Adult spelling errors are either typological or orthographical,
and many different types of error detection and error correction methods have been
developed. Although these studies have been done on English generally, the same
techniques can be applied specifically to medical information retrieval. However, the
challenge is that medical terminology or jargon often involves difficult, non-English
words. From the usability study it was seen that users often spell phonetically and
often use a scenario-based search rather than a keyword search.
Current spelling aid systems which are built on a phonetic basis might have high
recall but are low in precision. Combinations of phonetic-based systems with a
knowledge-based search are seen to improve searches by 5-10% (Liu & Chu, 2005).
What is more, Liu and Chu suggest that user queries are often scenario-based rather
than keyword based.
Pages:
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808