Towards Seamless Knowledge Integrating Public Sector Portals in

Towards Seamless Knowledge Integrating Public Sector Portals in Norway Topic Maps Published Subjects TMRAP Steve Pepper Chief Strategy Officer, Ontopia pepper@ontopia. net © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

What is Topic Maps? • Topic Maps is an ISO standard for Knowledge Integration • It is the only international standard for Knowledge Integration • But the more important question is… © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

What are Topic Maps used for? • That’s like asking: What are relational databases used for? • The answer is: A whole number of things, including (but not limited to): – – – Organizing large bodies of information Capturing corporate memory Representing complex rules and processes Supporting concept-based e. Learning Enabling Enterprise Knowledge Integration (EKI) • But in particular… • Any or all of the above, in combination! • Topic Maps lets you achieve Seamless Knowledge © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Seamless Knowledge • General business problem addressed by Topic Maps: – The disconnectedness of Information and Knowledge • Seamless Knowledge – A term coined within the Topic Maps community to describe the business benefits of applying Topic Maps • There is growing awareness of the scale of this problem: – Increased talk about “metadata”, “taxonomies”, “ontologies”, and “semantics” – The META Group talks of a “near-impending crisis” – What people are looking for is knowledge integration – i. e. , Seamless Knowledge – Topic Maps offers a standards-based solution • Seamless Knowledge is not the same as the Semantic Web – But there is some overlap and even more potential synergy © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Semantic Portals • One of many applications of Topic Maps – Topic Maps is an ideal model for portals and other forms of web-based information delivery • The basic concept is to have the topic map drive the portal – Not just a navigational layer on top of something else – The very structure of the portal is a topic map – All content is organized around topics (“subject-centric organization”) • Each page represents a topic (we call this a “Topic Page”) – Topics act as points of collocation – They provide a “one-stop shop” for everything that is known about a particular subject • Navigating the portal == Navigating the topic map – Associations provide very intuitive navigation (“As we may think”) © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

A Topic Page (multiple) types the current topic multiple names multiple typed occurrences multiple

Architecture of a Topic Maps Portal topic map data and documents © 2004 Ontopia

current topic occurrences associations

the current topic multiple names multiple associations multiple occurrences

The Rise and Rise of Semantic Portals in Norway • In Norway, this concept has been put into practice on a scale that is now verging on the industrial… – There are over a dozen topic map-driven portals in production – More are on the way… • And while the rest of the world is asking questions like – “Metadata? ” “Taxonomies? ” “Ontologies? ” • …in Norway, customers are saying “Topic Maps!” – – • Rf. Ps regularly specify Topic Maps as a requirement Headhunters are looking for Topic Maps experts 120 people attended the last Topic Maps Congress (Norway: pop. 4 million) Topic Maps are quickly moving from “early adopter” to “early majority” How did this situation come about? – The presence of Ontopia was important, but not enough on its own – We needed a high visibility success story as well… © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

The ITU Story (in brief) • Once upon a time, not long ago (in late 2000), … • … the Network for IT Research and Competence in Education (ITU) was planning a new web site • They had rather special requirements… – “Relationships between objects and various groups of objects offer users multiple paths to the same content and stimulate cross-site content exploration. ” – “Visualisation of this network is supposed to give the user a conceptual model of the network, and give a feeling of being in a ‘relational space’. ” • The consultant leading the project was Stian Danenbarger • At exactly the same time, XTM 1. 0 was announced: – “A standardized notation … used to define topics, and the relationships between topics. . . A topic map defines a multidimensional topic space (in which) locations are topics… relationships […] define the path from one topic to another. ” – A light bulb went on for Stian… – Ontopia helped him build an Open Source web-based content management and publishing system that was entirely driven by topic maps, called ZTM (Zope Topic Maps) • … and ITU got the web site it was looking for: © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

more associations current topic occurrences associations

The success of ITU started a trend • ITU was “bleeding edge” in early 2001 – Stian calls it a “technical base jump – without a parachute” – Such adventures are not for the faint-hearted • Since then Topic Maps Portals have become a proven and well established technology – …at least in Norway. . . • ITU was followed by web sites for the Norwegian Research Council, the Norwegian Consumers Association and many others… – Some of these are based on ZTM – Others are based on other Topic Maps engines • At present there are over a dozen, with more on the way © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Some Topic Maps Portals in Norway • In production • Under development • http: //www. itu. no http: //www. luna. itu. no (Ministry of Education) • Skatteetaten (Tax Office) • http: //www. forskning. no http: //www. nysgjerrigper. no (Research Council of Norway) • Statsministerens kontor (Office of the Prime Minister) • Statistisk Sentralbyrå (Central Bureau of Statistics) • IFE/Halden (Nuclear Reactor Project) • etc. • • http: //forbrukerportalen. no (Consumers Association) http: //www. skifte. no (Norwegian Defence) • http: //www. hoyre. no++ (Norwegian Conservative Party) • http: //matportalen. no (Ministry of Agriculture) • http: //www. udi. no (Ministry of Justice) • http: //www. kulturnett. no (Ministry of Culture) © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Towards Seamless Knowledge • As the number of portals multiplies, the amount of overlap increases… • Take these three portals as an example: • forskning. no (Research Council web site aimed at young adults) • forbrukerportalen. no (Public site of the Norwegian Consumer Association) • matportalen. no (Biosecurity portal of the Department of Agriculture) © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Genetically modified food at forskning. no

Genetically modified food at Forbukerrådet

Genetically modified foodstuffs at Matportalen

Three Topic Maps Portals – One Common Subject one “virtual portal” with seamless navigation

Towards Seamless Knowledge • Very little is required for these portals to achieve a simple but effective form of Seamless Knowledge • They have already achieved subject-centric organization of their content – Without this, Seamless Knowledge is beyond reach • From a technical viewpoint, only two additional pieces are required to complete the puzzle: #1 An identity mechanism – To make it possible to know when their subjects are the same #2 An exchange protocol – To enable information to be requested and exchanged automatically • (There must also be a real desire to share information, but that’s a political matter) © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Piece #1: The Identity Mechanism • Simply put: – How can we know that “genetically modified food” is the same as “genetically modified foodstuffs” (or “GM food”, or “genmodifisert mat”, for that matter)? • One thing is certain: Basing this on names won’t work – Synonyms, homonyms and polysemy make names a minefield – In any case we would like to multilingual knowledge integration • What is needed is nothing more or less than unique, “global” identifiers for all subjects of common interest • An impossible task? • Not if we go about it the right way… • In fact, the solution already exists in the form of a mechanism developed as part of the Topic Maps standard… • That mechanism is called Published Subjects © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

So what is Published Subjects? • An open, distributed mechanism for assigning unique, global identifiers to arbitrary subjects • Originally conceived as part of the Topic Maps effort, but applicability is far more general • The mechanism is based on using URLs as identifiers – e. g. Ibsen Museum in Oslo: http: //psi. kulturnett. no/museum/ibsen-museet • Nothing very special about that…, except that • The Published Subjects mechanism has two interesting characteristics: – It is two-sided – it works for both computers and humans – It works from the bottom up – not from the top down • Both of these are critically important… © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

For computers AND humans • Clearly any mechanism has to work for computers – In one sense, that is the whole point: – To make it possible for computers to decide when subjects A and B are the same – Only then can information be connected correctly • Computers can simply compare URLs • • • http: //psi. kulturnett. no/museum/ibsen-museet + http: //psi. kulturnett. no/museum/ibsen-museet = same subject http: //psi. kulturnett. no/museum/ibsen-museet + http: //psi. kulturnett. no/museum/ibsen-huset = not (demonstrably) the same subject But the mechanism must also work for humans – Because it is humans who (in the final analysis) actually assign the identifiers when structuring or classifying their information – A human needs to know exactly which is represented by a URL such as • • • http: //psi. kulturnett. no/museum/ibsen-museet (Is it the Ibsen-museum in Oslo or the Henrik Ibsen Museum i Skien? ) With PSIs, this can be done quite simply by clicking on the URL – The result is a document that provides some suitably unambiguous, humaninterpretable indication of the identity of the subject – We call this document a subject indicator © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

The dual nature of Published Subjects • The issue of identity is two-sided – It involves both humans and computers • The Published Subjects mechanism is similarly two-sided • The dual aspects are – a subject identifier (URL) used by computers – a subject indicator (document) intended for humans • The identifier is the address of the indicator • To understand what the identifier is intended to identify, simply click on it! • What could be simpler? © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

For computers AND humans A subject is identified via a URL subject • The URL is called a subject identifier topic Ibsen Museum http: //psi. kulturnett. no/museum/ibsen-museet • The URL is the address of a document • That document provides a human-interpretable indication of the identity of the subject • The document is called a subject indicator http: //psi. kulturnett. no/museum/ibsen-museet Computers use the identifier Simple comparison of string values: Identical values mean that the subject is the same Ibsen Museum located in the apartment in Arbiens gate in Oslo where the playwright Henrik Ibsen lived for the last 11 years of his life, from 1895 until 1906. • Humans use the indicator By inspecting the document one can be sure that the identifier does not refer to, say, the Henrik Ibsen Museum in Skien subject indicator © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Life, the Universe and Everything The concept of Subject Indicators in Topic Maps subject The identity of most subjects can only be established indirectly – Because it is a resource, a subject indicator has an address, even though the subject that it is indicating does not l uc ra /p subject indicator ps i. o nt op ia. Subject indicators and subject identifiers are the two sides of the human-computer dichotomy : // • subject identifier pe – Computers can use the address of the subject indicator to establish identity These are called subject identifiers ne t/o – The Computer Domain tp • Giacomo Puccini, Italian composer, b. Lucca 22 nd Dec 1858, d. Brussels, 29 th Nov 1924. Best known for his operas, of which Tosca is the most. . h tm – An information resource (like a definition or a picture) can provide some kind of indication of the subject’s identity to a human Such a resource is called a subject indicator A topic may have multiple subject indicators ci ni – ht • Puccini topic © 2004 Ontopia AS XML 2004, Washington DC The Topic Map Domain http: //www. ontopia. net/

Ramming the point home: Another diagram! The abbreviation PSI means published subject indicator, but could equally well mean published subject identifier © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

From the bottom up – open and democratic • Earlier top down attempts to create global identifiers have largely failed – or at least only met with moderate success – For example, URNs have been around for jonks and yet there are still only 17 official URN namespaces • • • Perhaps requiring a registration authority is too bureaucratic? Perhaps the inability to resolve URNs easily makes them difficult to use? Published Subjects uses the opposite approach… – Anyone can create a PSI (Published Subject Indicator) – The process is bottom up – open and anarchic – just like the Web itself • Survival of the most trusted – An evolutionary, darwinistisk process – The more authoritative, trusted, and respected the “publisher”, the more likely its identifiers will achieve widespread adoption • Emergence of de facto standards based on trust – The key parameter is confidence in the stability and longevity of the PSI © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Piece #2: The Exchange Protocol Hi! Do you know the subject “genetically modified food”? * * The actual question was: Is the subject http: //psi. forskning. no/food/gm-food known in your system? Sure. My URL is: http: //matportalen. no/Matp ortalen/Emner/gmo http: //matportalen. no/Matportalen/Emner/gmo Portal A: forskning. no This scenario is Level 1 of TMRAP knowledge integration. Portal B: Matportalen © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

TMRAP (Topic Maps Remote Access Protocol) • Abstract protocol for getting information from remote repositories – The protocol has an HTTP REST binding – A SOAP binding would be easy to do • Any repository can support TMRAP – For topic map applications support TMRAP is very easy – For other applications it’s less easy, but the benefit is that legacy applications can be integrated • The OKS currently contains a prototype implementation – Used to implement the Vizigator applet – Also used for the Omnigator Rap demo • For a short introduction to TMRAP: – http: //www. jtc 1 sc 34. org/repository/0507. htm • Some related work: – RDF Net API: http: //www. w 3. org/Submission/2003/SUBM-rdf-netapi-20031002/ – SNAPI: http: //sourceforge. net/projects/snapi © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

The Omnigator Rap Demo (Part 1: VISIT) • Two Omnigators are running on this machine – Different browsers (Opera and Internet Explorer) – Different skins (Ontopia National Colours and Vive Québec) – Different names pepper poivre – Different TMs (Italian Opera and Various Geographical TMs) • They are aware of each other’s existence • Their support for TMRAP is turned on © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

VISIT: Some Considerations • The functionality is deceptively simple, yet potential very powerful – From the user’s point of view the VISIT links might have been hand-coded (there is no visible difference) – The cool thing is that they are generated entirely automatically – This is spontaneous knowledge federation in practice!! • Think about it a bit: – Having multiple Omnigators rapping together is already fairly cool – In fact, any application built with the Ontopia Knowledge Suite can now join in the fun – And more importantly: – So can any application at all – whether or not it is based on Topic Maps – The only prerequisites are: • Subject-centric organization (i. e. , some concept of Topic Pages) • Use of Published Subjects (for the purpose of subject identification) • Support for TMRAP (in order to send and respond to requests) © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

VISIT: More Considerations • How useful is it really? • Isn’t it a little simple-minded? • For many of our customers it is sufficient as a first step – The Norwegian Research Council and the Norwegian Consumers’ Association want to be able to link to each other in this way – The VISIT paradigm enables them to retain their own branding – At the same time, they offer their users an extremely valuable service • TMRAP is already being implemented in ZTM – When done, not only will the Research Council and the Consumers’ Association be able to rap together… – …any Omnigator user will also be able to rap with them! • And remember: – This game can be played by any solution that uses some kind of subject-centric organization and PSIs © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

The Omnigator Rap Demo (Part 2: GET) • But we can go a step further with relatively little effort • Remember: Topic Maps are designed for merging … – … so we can exchange not only Topic Page URLs, – but also fragments of content in topic map form • We are calling those fragments topic maplets • TMRAP also supports exchanging maplets © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Piece #3: Topic Maplets (XTM fragments) Hi! What do you know about “genetically modified food”? * * The actual question was: What information do have about http: //psi. forskning. no/food/gm-food in your system? Oh, this and that. Here you are. Be my guest! Portal A: forskning. no This scenario is Level 2 of TMRAP knowledge integration. Portal B: Matportalen © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

GET: Some Considerations • The functionality is even more powerful… – The seamlessness factor is much greater – In fact we have “dumbed it down” in this Omnigator implementation in order to be able to show what is going on: The GET functionality could be activated automatically • Application areas are slightly different: – Useful when seamlessness is more important and branding issues less important • E. g. , within a corporate environment – Opens up the possibility of totally individualized “portals” • Topic Maplets – Raises some interesting technical issues – The most important is deciding exactly what the fragment should contain – TMQL (Topic Maps Query Language) will provide greater flexibility © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

The Building Blocks of Seamless Knowledge • Topic Maps – Semantically structured data that can be “viewed as topic maps” • • Without this, Seamless Knowledge is beyond reach By the way, this includes RDF, Relational DBs, XML and more – Already here • Published Subjects – The Semantic Superhighway – Globally unique identifiers for arbitrary subjects – Already here • Topic Maps Remote Access Protocol (TMRAP) – Protocol for requesting and delivering Topic Page URIs and Topic Maplets – Already here • Topic Maps Query Language (TMQL) – For more powerful and precise TMRAP requests – Watch this space (and use tolog in the meantime) © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Seamless Knowledge and the Semantic Web • Are they the same thing? – Not if you go by the vision of the Semantic Web articulated by Tim Berners-Lee (e. g. , in the famous Scientific American article) • – – • – Semantics are akin to knowledge … … and seamlessness implies the existence of something web-like … … so in a broader sense they do have a lot in common Certainly Semantic Web data (i. e. , RDF) will be easily reusable in the context of Seamless Knowledge (as will relational data and XML) The RDF and Topic Maps communities are currently working together to acheive interoperability at the data level (RDF/TM Interoperability Task Force) However, the TBL Semantic Web won’t be here for many years – • Most business users today don’t need AI and they don’t want to be restricted to the Web On the other hand, other people have other visions of the Semantic Web… In any case – – • In reality, that amounts to “AI on the Web” There is much research still to be done Seamless Knowledge is achievable today – Solving the problem of disconnected knowledge on a less ambitious scale| © 2004 Ontopia AS XML 2004, Washington DC http: //www. ontopia. net/

Conclusions • Topic Maps has almost “crossed the chasm” – at least in Norway – Web sites, Portals, E-learning, Knowledge Management, Enterprise Knowledge Integration, … • Seamless Knowledge is what Topic Maps is about – “Topic Maps” speaks only to the technology – CIOs are interested in business benefits and ROI • Published Subjects are the key to solving the identity issue – Anyone can create a PSI (Published Subject Indicator) – PSIs work for computers AND humans • TMRAP allows other data to be “viewed as” Topic Maps – Provided information can be made to look like a topic map, any legacy technology can play – The key is subject-centric organization of information • • • © 2004 Ontopia AS Without this, Seamless Knowledge is beyond reach Without this, Seamless Knowledge is beyond reach XML 2004, Washington DC http: //www. ontopia. net/