A fuzzy ontology for medical document retrieval ACSW Frontiers 2004: Dunedin, New Zealand 報告者...
-
Upload
sarah-richards -
Category
Documents
-
view
232 -
download
0
Transcript of A fuzzy ontology for medical document retrieval ACSW Frontiers 2004: Dunedin, New Zealand 報告者...
A fuzzy ontology for medical
document retrieval
ACSW Frontiers 2004: Dunedin, New Zealand
報告者 :楊仁杰 學 號 :9454604
outline1. Introduction
2. Fuzzy ontology
3. Membership values in a fuzzy ontology
4. Example of the use of fuzzy ontology
Introduction Searching for information can be an extremely complex task. Many searches performed are inadequate or cursory ex: the techniques of critical reflection
the different uses to which different groups ofmedical professionals use online tools, andhence the different types of resources required are
shown in the work
Introduction These approaches provide large amounts of data, b
ut do not really give any information on the intensions of the searcher, whether their search was successful
One way of overcoming this problem is the postulation of a “fuzzy ontology”.
Fuzzy ontology The fuzzy ontology is based around the
concept that each index term or object is related to every other term (or object) in the ontology , with a degree of membership assigned to that relationship based on fuzzy logic
fuzzy ontology = fuzzy logic + ontology
Fuzzy ontology ontology :
The use,reuse and sharing of information between ontologies, isespecially important in the medical field with the growthof evidence-based medicine(Moody and G. 1999)
Ontology members
“is a” 、 “ uniquely_mapped_to” and “developmental_form_of”.
Fuzzy ontologyfuzzy logic :
The fuzzy membership value μ is used for
the relationship between the term or object
in question where 0<μ<1, and μ
corresponds to a fuzzy membership relation
such as “strongly”,“partially”, “somewhat”,
” slightly”
Fuzzy ontology
A fuzzy ontology scheme.
Fuzzy ontology and collective searching Each user would have their own values
for the membership assigned to terms in the ontology,reflecting their likely information need and world view
Fuzzy ontology and collective searching
This method assigns a membership value for each potential relation for each term
(1)
(2)
Fuzzy ontology and collective searchingEx :
(L1, L2, L3) -> (μ1=0.6, μ2=0.3, μ3=0.1)
L1’s membership value is decreased to 0.5
(L1, L2, L3) -> (μ1=0.5, μ2=0.375, μ3=0.125)
Fuzzy ontology
Membership values in a fuzzy ontology
Fuzzy ontology
5 Parent terms + 19 Child terms
Membership values in a fuzzy ontology
Membership functon for fuzzy ontology
Membership values in a fuzzy ontology
Manual assignment of membership values :
The user can then select these terms according to the degree of relatedness to the original query term. There are boxes for “Opposite”, “Not Related”, “Slightly Related”, “Moderately Related” and “Strongly Related”. A value for “Usefulness” of the document is also recorded.
Membership values in a fuzzy ontologyManual assignment of membership values
A value for “Usefulness” of the document. “Opposite”
A value for “Useless” of the document.
Ex : “Cold” the term “Frigid”
“Opposite”
“Cold” the term “torrid”
Membership values in a fuzzy ontology
Automatic Membership value assignment :
Queries run against both GOOGLE (
www.google.com) and PubMed (pubmed.gov)
each term that exists in multiple locations, is given an equal and proportionate membership value
Membership values in a fuzzy ontology
weighted 3 2 1
PubMed keyword section
Title abstract
Google meta tags Headings main body
local terms (Li)
all terms discovered (Ai)
Example of the use of fuzzy ontologyAND - conjunction
The fuzzy “AND” operator used in this thesis is the min operator.
ex : “ Head” and “Nose” (0.7,0.9)membership
value
a query : “Head AND Nose” ->”Head”
Example of the use of fuzzy ontologyOR - disjunction
The fuzzy “OR” operator used in this thesis is the max operator.
ex : “ Head” and “Nose” (0.7,0.9)membership
value
a query : “Head OR Nose” ->”NOSE”
簡 報 完 畢簡 報 完 畢敬 請 指 導敬 請 指 導