Curriculum Encoders Group
Ontology Modifying Etiquette
Added or modified by Maxim Hendriks on Jan-28-2011
We need a written down version of rules on how we structure the ontology. I want to gather these here for now.
Rules on Identifiers
Terminology
Ontology nodes have an identifier and several names. A node has a unique identifier and several names (common names, uncommon names, rare names and false friends, to be precise) for various languages. For example the node with identifier "Lattice_Point" has an English common name "Lattice Point" and a Dutch common name "Roosterpunt".Rules
Rules on Identifiers
- The identifier should be in English whenever this is possible. This will make the node identifiers easier to understand for most people, and will thus not lead to a curriculum encoder adding a new topic because they think it's not there, while someone made a node for the topic already with a Czech identifier. We want to avoid creating several nodes for the same competency or topic for at least two reasons: 1) the ontology stays smaller, without lots of duplication (27 nodes for "triangle") and 2) it is actually the whole idea of the ontology, because having 1 node with several names in different languages relates these expression, thus increasing exchange between the various language regions. If you are thinking of giving a node a French identifier, think again five times, then ask all other curriculum encoders about an English name, then look in the dictionary, and finally create it in English.
- The identifier should start with a capital letter. A bit arbitrary but good for uniformity, so that if some other software developer wants to use the ontology, it's not impossible to guess node identifiers.
- For the same reason of uniformity, if an identifier is an amalgam of multiple words as written in English, then these should be separated by an underscore '_', and each word should start with a capital letter. So "Lattice_Point", and not "lattice_point", "LatticePoint" or "Latticepoint".
- Identifiers are not allowed to contain accents, only the characters A-Z, a-z, digits 0-9 and "_".
- If you want to add a subtopic to a topic that is now an individual, 1) delete the individual, 2) create a class with exactly the same identifier, 3) create an individual with that identifier + "_r" as a member of the class, and 4) create your new individual in the class as well. Note that your new individual should not have an identifier ending in "_r". So if you were to add the concept 'right triangle', where now there is only the individual "Triangle", delete individual "Triangle", add a class "Triangle" and an individual "Triangle_r" which is a member of this class, then finally create an individual "Right_Triangle" as a member.
- Classes and individuals describing objects are named in singular. So "Point", not "Points".
- Relationships are named using adjectives. So "Symmetric" is a relationship, "Symmetry" is an operation or a property of figures.






