National Data Standardization: A Place for Topic Maps?


Published on

This paper deals with national data standardization efforts in Denmark and discusses the role Topic Maps – and topic maps – may play in a new standardization strategy currently being considered by the Danish National IT and Telecom Agency. The strategy entails a paradigm shift from syntactic data standards based on XML schema to a more semantically based approach involving, among other things, the development, publishing and sharing of so-called definitions. The paper accounts for the historical, political and technical context of the strategy pointing out some of the opportunities and constraints this context poses for the introduction and application of Topic Maps as a recognized “data standardization standard” in Denmark.

Published in: Technology, Education
  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide

National Data Standardization: A Place for Topic Maps?

  1. 1. National Data Standardization: A Place for TopicMaps?<br />Lars Johnsen <br />
  2. 2. ”There is something rotten in the state of Denmark”- Shakespeare (1601) <br />2009: Challenges =&gt; modernization of the public sector =&gt; digitalization =&gt; interoperability<br />
  3. 3. To beor not to be – standardizing *<br />2001: It standardizationeffortstakingoff<br />2006: B103 – Parliament voting for open standards in Danish e-goverment<br />2008: Seven sets of open standards made compulsoryincluding web standards, document standards and data standards<br />Where is TopicMaps …?<br />* … that’sone of the questions<br />
  4. 4. Data standardization in Denmark<br />Background and status<br />New strategybeingimplemented by ITST (the national agency for It and Telecom)<br />Technologicalinfrastructure<br />Paradigmshift from syntax to semanticsorevensubject-centricthinking …<br />Why not TM? <br />Agenda<br />
  5. 5. Data standards - OIOXML<br />OIOXML schemas<br />Domain schemas<br />Corecomponents<br />
  6. 6. Problems with OIOXML schemas<br />Content<br />Little documentation<br />Whatlegislation?<br />Whatcontexts?<br />Whatprojects, web services, etc.<br />Findability and usability<br />Uploading<br />Organizing<br />Viewing<br />Concepts<br />No real semantics:<br />&lt;MaritalStatusCode&gt;<br />married<br />divorced<br />widow<br />registeredpartnership<br />abolition of partnership<br />longestliving partner<br />deceased<br />unmarried<br />&lt;/MaritalStatusCode&gt;<br />What do youread, my lord? <br />Words, words, words<br />
  7. 7. A new strategy in the pipeline<br />Taxonomies<br />Concept systems<br />Semantic definitions<br />Message definitions<br />Data definitions<br />OIOXML schemas<br />WDSL files<br />
  8. 8. Concept systems: buildingonexistingmethodologies<br />
  9. 9. From concept systems to taxonomies and thesauri<br />FORM – a reference model of public services in Denmark <br />Social services<br />Benefits<br />Individualbenefits<br />Action<br />Allocation of housingbenefit<br />Object<br />
  10. 10. In the future: semantic definitions as pivots<br />&lt;SemanticDefinitiondefinitionIdentifier = ”…” systemIdentifier = ”…” versionIdentifier = ”…”&gt;<br /> &lt;LocalizationlanguageCode =”DA”&gt;<br /> &lt;Term roleCode = ”preferred”&gt;…&lt;/Term&gt;<br /> &lt;Term roleCode = ”admitted”&gt;…&lt;/term&gt;<br /> …<br /> &lt;DescriptionroleCode = ”definition”&gt;…&lt;/Description&gt;<br /> &lt;/Localization&gt;<br /> …<br />&lt;/SemanticDefinition&gt;<br />
  11. 11. Semantic<br />definition<br />From semantics to syntax<br />Semantic<br />definition<br />client<br />has<br />client status<br />is a<br />Data definition<br />Client<br />Client_status: custody, …<br />Semantic<br />definition<br />custody<br />“Though this be madness, yet there is method in it”<br />OIOXML<br />&lt;ClientStatusCode&gt;<br /> &lt;enumerationvalue =”custody” /&gt;<br /> …<br />&lt;/ClientStatusCode&gt;<br />
  12. 12. Conceptuallayer<br />Concept systems and taxonomies = TAO <br />topics/subjects, properties, associations, roles<br />Semantic definition = PSI<br />names, descriptions, explanations, etc. of subjects (in oneor more languages) <br />Contentorresourcelayer<br />OIOXML schemas, data definitions, wsdl files = occurrences/addressablesubjects<br />Embedded OIO schemas = part-whole associations betweenaddressablesubjects, etc.<br />TM model, really<br />
  13. 13. So, why not simplyadopt TM as a unifying model or standardfor representing and organizing ”data standards data” in Denmark …? <br />Instead an entirely new OIOXML language is beingdeveloped …<br />Thisquestionbecomeseven more relevant …<br />
  14. 14. Web 2.0 platform (<br />Information architecturecenteredongroups<br />Functionality for news, debates, usernetworks, etc.<br />Tagging as primaryorganizationvehicle<br />REST API<br />Resources, representations and structured URL’s<br />Finding, publishing and sharing ..<br />Prediction: a direneed for an organizational/navigationaloverlayintegratingdistributedconceptual information, resources and metadata<br />
  15. 15. Why not adopt TM as a unifyingtechnology for organizing and integrating data standards (and elsewhere)? <br />
  16. 16.
  17. 17. TopicMaps as a technologyworking ”covertly” behind the scenes ortopicmaps as accessible information products in theirown right? <br />
  18. 18. Topicmaps as<br />OIO topicmaps – trulyopen information sets about the ”state of Denmark”<br />
  19. 19. The vision of OIO TopicMaps …<br />Public institutions and authorities<br />Public services (e.g. FORM)<br />Integration<br />Public web services<br />Coreconcepts<br />Laws and guidelines<br />Workflows<br />
  20. 20. &lt;topicMap version=&quot;2.0&quot; xmlns=&quot;;&gt;<br />&lt;topic id=&quot;enke&quot;&gt;<br />&lt;subjectIdentifierhref=&quot;; /&gt;<br />&lt;subjectIdentifierhref=&quot;; /&gt;<br />&lt;instanceOf&gt;<br />&lt;topicRefhref=&quot;#OIOconcept&quot; /&gt;<br />&lt;/instanceOf&gt;<br />&lt;name&gt;<br />&lt;scope&gt;<br />&lt;topicRefhref=&quot;#DA&quot; /&gt;<br />&lt;/scope&gt;<br />&lt;value&gt;enke&lt;/value&gt;<br />&lt;/name&gt;<br />&lt;occurrence&gt;<br />&lt;type&gt;<br />&lt;topicRefhref=&quot;#OIOXML&quot; /&gt;<br />&lt;/type&gt;<br />&lt;resourceDatadatatype=&quot;;&gt;<br /> &lt;MaritalStatusCode&gt;widow&lt;/MaritalStatusCode&gt;<br />&lt;/resourceData&gt;<br />&lt;/occurrence&gt; <br />&lt;/topic&gt;<br />...<br />&lt;/topicMap&gt;<br />XTM 2.0 comes in handy …!<br />
  21. 21. General lack of knowledge<br />”TopicMaps is a metadata standard like Dublin Core”<br />”Conceptslikesubject, topic, identity and mergingaresimplytoohard to explain”<br />”Weare not havingthat RDF and OWL stuff, thereforewecannotadopt TM either”<br />The TM vocabularydoes not translateveryeasilyinto Danish<br />Why is TM not a national standard?<br />And of coursewe do not have a Lars Marius or Steve P …<br />
  22. 22. Emne - subjekt (topic/subject)<br />Identifikator - indikator (subjectidentifiervsindicator)<br />Adresse (locator)<br />Belæg – hændelse - forekomst (occurrence)<br />Sammenslutning – sammenlægning - fusion (merging)<br />Translating TM into Danish <br />
  23. 23. Play the standard card (”TM is an international ISO standard, youknow”)?<br />Create solutions! (chicken and egg problem) <br />RDF: friendorfoe?<br />The ”versatility problem”: promoting TM as Semantic Web technology, metadata format, orwhat? <br />So, whatshouldwe do? <br />PerhapsDigitaliser.dkwillsolve the problem for us …<br />”The rest is silence!”<br />