M edical c n I formati s Group i Indexes, Topic maps Indexes, Topic maps & the semantic & the semantic net net Dr Jeremy Rogers MRCGP Dr Jeremy Rogers MRCGP Clinical Research Fellow Clinical Research Fellow Medical Informatics Group Medical Informatics Group University of Manchester University of Manchester
17
Embed
MedicalcnIformatisGroup i Indexes, Topic maps & the semantic net Dr Jeremy Rogers MRCGP Clinical Research Fellow Medical Informatics Group University of.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Medical cnI formati sGroupi
Indexes, Topic maps Indexes, Topic maps
& the semantic & the semantic netnetDr Jeremy Rogers MRCGPDr Jeremy Rogers MRCGP
Clinical Research FellowClinical Research Fellow
Medical Informatics GroupMedical Informatics Group
University of ManchesterUniversity of Manchester
Medical cnI formati sGroupi
OutlineOutline
What’s an Index ?What’s a Topic Map ?
UsesStrengths & Weaknesses
Medical cnI formati sGroupi
What’s in an index ?What’s in an index ?La Bohème, 10, 70, 197-198, 326 Cavalleria Rusticana, 71, 203-204 The Girl of the Golden West, see La fanciulla del West Leoncavallo, Ruggiero
I Pagliacci, 71-72, 122, 247-249, 326 Madama Butterfly, 70-71, 234-236, 326 Manon Lescaut, 294 Mascagni, Pietro
different different typestypes of topic (the names of of topic (the names of operas are shown operas are shown in italicin italic))
different different typestypes of occurrence of occurrence (references to synopses are (references to synopses are in boldin bold))
synonyms for the same topicsynonyms for the same topic
links to associated topics (links to associated topics (see alsosee also))
associations between different topics associations between different topics (e.g. between a composer and his (e.g. between a composer and his works)works)
supertype and subtype informationsupertype and subtype information
Medical cnI formati sGroupi
What’s a topic map ?What’s a topic map ?
• Very similar to an index except…• Implicit information made explicit
– Not represented in typographical conventions
• Consistent use of topics within indexes– Consistent use between indexes ?
Medical cnI formati sGroupi
What is a topic map ?What is a topic map ?www.topicmaps.org
Strengths of Topic Maps…Strengths of Topic Maps…• Good for large or dynamic
information sources– E.g. ‘2001 Tax Products’
CDROM from US IRS– E.g. www.quid.fr
• Better querying than a static index– ‘Find all composers who have
written an opera that was not first performed in Italy but was based on a work originally written by Shakespeare’
• Making it possible/easier to merge knowledge resources– allegedly
Medical cnI formati sGroupi
……and limitationsand limitations• Not so good for very large
information sources– E.g. The entire world wide web
(= ‘The Semantic Web’)– The ontology problem
• Different topics used differently • at different times & by different authors
• No inferencing (yet)– E.g. IF Aria sung-in Act I
AND Act I part-of ToscaTHEN Aria sung-in Tosca
• Limitations in what can be said– E.g. Child must have exactly one mother– E.g. Mother of a baby kangaroo must
also be a kangaroo
Medical cnI formati sGroupi
Current DevelopmentsCurrent Developments
• Standardising the query language– ISO/IEC 18048 TMQL
• Standardising the constraint language– ISO/IEC 19756 TMCL (based on OWL
?)• “All persons must be born somewhere”• “A person may have died somewhere”
• Experiments with inferencing
Medical cnI formati sGroupi
Topic maps vs. semantic Topic maps vs. semantic netsnets
A Semantic Net is any data structure network with nodes representing objects, classes or concepts, and the arcs representing relations between them (e.g.
is-a, part-of etc)
•Therefore, all topic maps are a kind of semantic net
•But not all semantic nets are topic maps– Because richer semantic nets may include e.g. cardinality– Not all semantic nets need external references
•Other kinds of semantic nets:– Mind maps– Entity Relation Diagrams & UML (databases)– Description Logics (ontologies)
Medical cnI formati sGroupi
SummarySummary• Topic Maps formalise back of book
indexing– Especially useful for electronic age– Makes the index machine-readable
• Aspirations for more– Usual plans for global domination
• Relatively young– Details still being finalised– Low uptake so far– But key exemplars