DrDC Game Technical Matters: Use of OWL-DL for Game Ontology



DrDC Game Technical Matters
The DrDC game is to be played on the WWW. Like all games it has rules and a playing field which we call the magic circle.

Magic Circle:
The WWW itself provides the basic ground for the playing. But we need to restrict it in some way in order that players will know when they are on the playing field or not. In other words we need to define the magic circle of play. This is a matter of current research. Currently, rules are used to define the magic circle.

Rules:
Let us imagine that Google is the only search engine permitted? This is like saying that a ball can only be kicked with the feet and not touched by the hands in the game of football. Searching with Google certainly limits the type of play but not much.
Let us imagine that any player is permitted to search in their own language as well as the lingua franca (English). For definiteness, suppose that two players agree to use Gaeilge and Bulgarian, in addition to English. Then to a certain extent use of these languages also restricts the magic circle. Some of the rules of play have already been introduced earlier in this issue of the Journal.

Computability:
Since the DrDC game lives in a digital world on the WWW it is natural that all relevant matters concerning computability should be addressed. For the past year we have been focussing on the transcription or recording of games. We have assumed that the played game will be available to anyone on the WWW in a form that will enable them to replay it in order to learn how such a game is played or to gain further experience, much as one might do to learn or improve in the playing of chess. But we need to do more for computability. In order that search engines (such as Google) can make “cultural” connections independently of human intervention our DrDC games must conform to the Semantic Web Ontology Language: OWL. [http://www.w3.org/2004/OWL/]. For practical research purposes we are using a free Ontology editor: Protégé and the OWL plugin, available from Stanford University [http://protege.stanford.edu/plugins/owl/].
OWL-DL
For the DrDC Ontology we deliberately restrict ourselves to that OWL based on Description Logics (DL), the latter being “a family of knowledge representation (KR) formalisms that represent the knowledge of an application domain (the “world”) by first defining the relevant concepts of the domain (its terminology), and then using these concepts to specify properties of objects and individuals occurring in the domain (the world description).” (Baader et al. 2003, 43).
A simple example or two of OWL-DL will help:
Ō hasKeyword. Butterfly -- individual games having butterfly as a keyword
ÅÕ hasKeyword. Butterfly -- individual games all of which have butterfly as a keyword.
The concept of Butterfly Game may then be defined by the concept expresion
Ō hasKeyword. Butterfly hasKeyword. Butterfly,
which denotes the set of individual games having at least one filler of the role hasKeyword belonging to the concept Butterfly and moreover, every filler of the role hasKeyword must be a Butterfly.
Let us now put together the concepts of URL and Image, both of which are key elements of a DrDC game. The corresponding roles are hasURL and hasImage, respectively. Consider the concept defined by the DL expression (≥2 hasImage) (≤ 3 hasURL). This represents the concept of “individual games having at least 2 images and at most 3 URLs.” It is clear to see that the Ohrid Game is such a game. In the actual Ontology we are constructing we use a very general role called hasKey. Subroles of this are hasKeyword, hasKeyphrase, and hasKeyimage. We can imagine that the Key element to a game might be an image rather than a word.
Corresponding to each role, such as hasKeyword, there is an inverse role, isKeywordOf. Details of the OWL-DL Ontology for the DrDC game will eventually be released into the public domain at
http://homepage.mac.com/micheal1/iblog/B1888672450/index.html

Posted: Thu - September 1, 2005 at 03:23 p.m.          


©