Topic Maps

Topic Maps: An Introduction 

Bob Bater, InfoPlex Associates

Bob provided a good overview on topics maps to help to ground the participants in the basics of topic maps including their origin, some key concepts, how they support knowledge maps, and how they can lead to knowledge discovery.
Bob walked through the evolution of SGML to XTM and XML. 

He described the index as a 2d model for finding relationships while a topic map represents a 3d model with a greater ability to show relationships between topic items. The associations between items is what makes the terms better references, as well as the ability to link in actual references to each term.

He described how Topic + Association + Role + Scope = Precision

He then described a couple of examples to help bring it home. He walked through an example to show how it relates to a knowledge map and even built an activity system ontology. This led to the development of a competence ontology. He then showed a couple of screen shots from Omnigator to show what it provides and how it supports the development of topic maps.

Topic Maps: Compelling Applications of a Simple Idea
Steve Newcomb, Coolheads Consulting

Steve started by describing the most important idea of topic maps: one virtual location for all things relating to a topic. It can be accessed from many different places but the idea is that once you get there, everything you need is there.

He introduced the concept that topic maps act as “wormholes” to connect different universes. The different universes are the collection of terms that one group may have and want to connect to another group’s collection of terms. The term collections are linked together by a common topic. The topic maps are means by which you can traverse the universes and find content that is related yet given a different set of metadata. The relationships and associations between topics are what provide the appropriate context for focused searches.

He then launched into a series of examples of topic maps in current use, including the IRS, the US Navy’s Office of Naval Intelligence, the Atlas development process,  municipal portals, US Department of Energy, Dutch Tax and Customs, and the German Resource Center for Genome Research. 

He concluded his talk by encouraging people to attend the topic maps research and applications conference (http://www.tmra.de).

Are Topic Maps the Final Piece of the Puzzle?
Marti Heyman, Deloitte Touche Tohmatsu

Marti presented the challenge of findability of content they have at Deloitte and how they are experimenting with topic maps to find a solution. They started by helping enable learning and knowledge discovery with advanced search and browsing techniques. Their next objective is to build knowledge maps utilizing semantic techniques to enable further associations and a rich enterprise-wide thesaurus.

Marti showed a screenshot of what their search capability allows them to do and then described how they need to work on making their taxonomy easier to navigate. They do not provide any tools to help a user understand how their taxonomy is structured or how the different nodes relate to each other. Marti explained how they are using omnigator to explore whether the development of topic maps will support their overall goal to enable the user to better traverse their taxonomy and act like a knowledge map. Once the user navigates to the appropriate node, they can see the content associated with that node and its relationship to other nodes.

Some of the concerns that they have is that the software is beta and may not be robust enough to handle the demand the system will place on it. Cost and maintenance are additional concerns that will need to be mitigated.

Service Oriented Architecture in the presence of Information Structure
Paul Prueitt, Ontologystream

Paul presented his premise that the goal is to measure in real time the exchange of information and that information occurs within a community. He then proceeded to set the stage by providing his definitions of different knowledge elicitation techniques, taxonomy and controlled vocabulary, OWL business ontology, and topic maps. In a series of pictures, he proceeded to show how OWL is almost perfect but that it needs to have human interaction to present the appropriate context. By establishing a semantic interpretation environment, it facilitates the connection of the information structure to a community of practice to establish the contextual relationships necessary for meaningful knowledge transfer.