Merging

In Topic Maps, a topic represents one, and only one subject. Several people can each work on a topic map and merge them afterwards. When two topic maps are merged, topics that represent the same subject should be combined into a single topic. The resulting merged topic contains the characteristics of both original topics combined. In short, merging multiple files results in a single topic containing the characteristics of both.

Suppose we have registered all Members of Parliament in a topic map, we should see 150 people. If we have a topic map of all MPs in the Balkenende IV cabinet and an older topic map from two years earlier, then there is obviously a difference. In the older topic map, we find Arend Jan Boekesteijn. Suppose that this would be the only difference and you would add the two lists of MPs together, we would get a list of 300 MPs without filtering, including 298 duplicate names and two single names: those of Boekesteijn and of his replacement (Mark Harbers).

With the merging character of topic maps and strong identification would we get 151 MPs, because doubles are merged. The power of merging is truly apparent by applying it yourself. With the Open Source software Ontopia you can get to work relatively easily, for example, to remove duplicates from your own address records.

Deduplication is obviously not enough. Therefore we want to ensure that for each cabinet and/or time period we record when a person is Member of Parliament. This can be done with Scope.

Scoping

Scope makes it possible to:

  • Express the subjectivity of knowledge
  • Look at a single source of knowledge from different perspectives
  • Provide personalized views for different user groups

With the help of scope we can ensure that the relation member-of is only valid in a certain time period or in a particular cabinet. In the figure below we show that A. J. Boekestijn was MP in the Balkenende IV cabinet until November 24, 2009.

Figure 2: Using Scope as an additional filter mechanism on an association

Figure 2: Using Scope as an additional filter mechanism on an association

With scope it is also possible to easily add claims about existing relationships. In the knowledge-layer, it is therefore possible to add a personal view to knowledge. This can be very useful in the following cases:

  • When different users have different requirements about knowledge;
  • To add a personal touch based on personal preferences, skills or security.

A particular bill may for example be labeled as very positive for a particular party or position, but maybe there are also some questionable or negative aspects of the same bill for the same party.

In a number of projects in which Topic Maps is applied, the example of scope shown above is used in a similar fashion: after the dissemination of several systems, the Topic Maps application offers a classification semi-automatically. In the Police case the system offers, for example: this person is probably a member of this group. The analyst then needs to confirm or reject the statement and possibly provide an explanatory supplement. In addition to the use of the above Topic Maps concepts, the so-called Topic Maps Query Language [i] is used for this application.

[i] An introduction to this language: http://www.xml.com/pub/a/2005/06/01/tmql.html

Pages: 1 2 3

Leave a Reply