The skeleton of the methodology is illustrated by the following diagram.
Figure 1. Skeleton of the Uschold and King methodology
1. Identify the objective: the objective of the ontology can be formulated as answers to the following questions:
What is the ontology built for?
What are its intended uses?
What is the intended user group?
At this stage we must determine questions of competence: Questions which must be spread by the ontology. Also, we identify relevant terms in the domain.
For example: developing a Travel Ontology whose objective is to build a consensual knowledge model on the travel domain, used by travel agencies, containing catalogs on means of accommodation and transport.
the terms: Places, types of places, Accommodation, types of accommodation, trains,…
2. Construction of the ontology
2.1 Capturing the ontology: Tasks to undertake
1) Identify key concepts and relationships. Three Strategies for Identifying Concepts:
- Top-Down: from the most abstract concept to the most specialized. This strategy risks imprecision. Example,
- Bottum-up: from the most specialized term to the most abstract. With this strategy, it is difficult to extract properties common to concepts. We risk producing concepts that are not important for the final ontology. Example,
-Middle-out: This is the most used. We start with the basic concepts. Example,
2) Write unambiguous texts for definitions of concepts and relationships
3) Identify terms to reference concepts and relationships
4) Agree on all of the above, creating a document containing the terms in natural language.
Example: Travel ontology
- Means of transport: is a class. Every means of transport has a starting point
- Bus: is a class. Is a kind of means of transport
- Local Bus: is a class. Is a Bus whose departure seat, destination seat and stops are in the same Location.
2.2 Coding of the ontology: This is the phase of implementing the conceptual structure by choosing a language. We can cite the following criteria to decide the choice of language:
- Outlook
- Conceptual distance
- The power of expression
- Alignment with standards
- Support for methods and guidelines
- Formal semantics to check consistency.
2.3 Integration of existing ontologies: This task can be during the capture and coding processes. For example, entities from the Standard-units ontology on measurement units can be imported into the travel ontology to measure the lengths of travel routes. In general, we can use tools that help us identify
ontologies to reuse.
2.4 Evaluation: the evaluation of the ontology can be formulated as answers to the following questions:
- Does the ontology meet the needs?
-Does the ontology answer the skills questions?
2.5 Documentation:
- guidelines for documentation
- Very important for sharing knowledge
- All assumptions must be documented.
