Topic Maps Explained

Topic Maps is an international standard for storing knowledge and information on the basis of concepts and relationships. The standardization of Topic Maps is provided by a working group [i] of ISO (International Organization for Standardization). Originally, the standard was prepared [ii] in the early 90s to create a master index for documentation in which different indexes were mixed. There was already a great need at that time to unlock multiple sources and to create one overview. In 2001, there was a Web standard called XML Topic Maps. Later on, the XML group of specialists worked on various other exchange formats based on the same data model. This data model remains the heart of the standard; it is the map in which aspects of a knowledge organization can be described in a computer-interpretable form.

The representation of the knowledge and information of an organization is shaped by following the specifications of the Topic Maps standard: it is a map of concepts and relationships and is referred to as a topic map [iii]. The base (the TAO [iv]) of Topic Maps consists of:

Topics: the main subjects;
Associations: relationships between topics;
Occurrences: ‘occurrences’ in the data, this can be facts, links to web pages, documents, fields in a database, etc.

In a topic map about Dutch politics, Pechtold (group chairman) and D66 (parliamentary group) can be topics, member of parliamentary group is a possible association between these two and the website of D66 can be an occurrence of the topic D66.

This is the basis from which you start working, but there are several other starting points to use to represent knowledge flexibly in the map. Topics are surrogates, or literally ‘proxies’ for topics that can be discussed. These topics can be any topics in the broadest sense of the word: for example, the dossier AOW is a topic and as said before, D66 is a topic, but also less tangible subjects such as Financial crisis or whether the commitment made by a minister about “Letter to the Chamber whether the current ICT infrastructure of the government is ready for Web 2.0” will be met before a certain date.

Besides the basic (Topics, Associations, Occurrences), there is concept of Types within the standard; types provide so-called class instance semantics, it indicates to which class or classes a topic belongs (of what it is an instance).

The following statements are in our example: Balkenende is a person, CDA is a party group;

Furthermore, we denote the following characteristics around topics:

  • Topics have identity. Balkenende for example, can be identified with his birth, his Social Security number or his email address. (more about identity follows).
  • Topics have characteristics:
    • zero or more names: name balkenende, but also JP is a nickname
    • zero or more occurrences
    • zero or more associations

[i] http://www.isotopicmaps.org
[ii] This origin took place at SOFABED = Standard Open Formal Architecture for Browsable Electronic Documents
[iii] Topic Maps (plural, uppercase) is used for the method; a topic map (singular, lowercase) is an information system built based on the method.
[iv] Paper by Steve Pepper : The Tao of Topic Maps, http://www.ontopia.net/topicmaps/materials/tao.html

Pages: 1 2 3

Leave a Reply