[Translation] what is an owl ontology?

ArticleDirectory

• 1.1 owl-Lite
• 1.2 OWL-DL
• 1.3 owl-full
• 3.1 individuals (individual)
• 3.2 Property)
• Class)

This article is a translation of chapter 3 of document a practical guide to building owl ontologies using the protege-owl plugin and CO-ODE tools edition 1.0, with pictures omitted. Proté Gé is an open source software entity editor developed by Stanford University. It has excellent design and many plug-ins. It is one of the most widely used ontology editors.

What is an owl ontology?

We use ontology to obtain the knowledge of a certain domain, describe the concepts of this domain, and the relationship between these concepts. Currently, there are many different ontology languages, each of which has its own merits. The latest W3C (World Wide Web Consortium) standard is owl. Like protégé, owl makes it possible to describe various concepts, while providing many other functions. It has richer operators, such as and, or and non-operators. It is based on a different logical model, which can better define concepts, A complex concept can be constructed from a simple concept. This model also allows you to use the inference engine (Reasoner) to check whether the statements and definitions in the ontology are consistent, you can also determine which concept is more suitable for you to maintain a correct Ontology. This is critical when multiple parent classes are allowed for a class.

First and third types of owl

Owl can be divided into three sub-languages: owl-lite, OWL-DL and owl-full, the main classification is based on their ability to express. Among them, owl-lite is the Child language with the weakest expression ability, owl-full has the strongest expression ability, while the OWL-DL expression ability is between them. We can think of OWL-DL as an extension of owl-lite, while owl-full is an extension of the OWL-DL.

1.1 owl-Lite

In terms of syntax, owl-lite is the simplest of the three. When the hierarchy of classes in your ontology is very simple and there are only simple constraints (constraint) it is suitable to describe the ontology. For example, when you need to port an existing Dictionary (thesauri) to another simple conceptual level, owl-lite can do well and quickly.

1.2 OWL-DL

Compared with owl-lite, OWL-DL has much more expressive power, which is based on description logics (the origin of dl ). The description logic is a deterministic variant of the first order logic (not necessarily accurate, the original decidable fragment), so it can be used for automatic reasoning, computers can then know the classification levels in the ontology and whether the concepts in the ontology are consistent.

1.3 owl-full

Owl-full is the most expressive language among the three sub-languages of owl. It is suitable for those who need very strong expressive ability, and do not care too much about Decidability) or when the computation is complete. However, because the expression capability is too strong, the ontology represented by owl-full cannot carry out automatic reasoning.

Ii. Which language should I use?

First, for more information about the differences between these three languages, see the Introduction to the owl Web Ontology Language. Although there are many factors to consider to decide which of them to use, here are some of the simplest and most commonly used principles.

• For owl-lite and OWL-DL, consider whether the simple construct child provided by owl-lite is sufficient to describe your ontology, if you use owl-lite, otherwise use owl-DL.
• For OWL-DL and owl-full, consider whether automatic reasoning is more important or highly expressive in your application (for example, whether meta-classes are needed for modeling, use OWL-DL, otherwise use owl-full.

The owl plug-in of proté Gé does not differentiate between the owl-lite and the OWL-DL ontology when editing, but you can choose to edit the ontology either in OWL-DL or in Owl-full mode in options.

Iii. Composition of the owl Ontology

The composition of the owl ontology is similar to that provided by prot é Gé. Basically, it is only called separately for the components. For example, owl has an individual (individual), a Property (property), and a class (class), while prot é Gé calls them instance and slot respectively) and class ).

3.1 individuals (individual)

Individual representatives (in the field) objects we are actually interested in. One important difference between prot é Gé and OWL is that owl does not use the unique naming assumption (unique name assumption, UNA ), that is to say, two different names can correspond to the same individual. For example, "Queen Elizabeth", "Queen", and "Elizabeth? Windsor "May all correspond to the same person. In Owl, you must specify whether the individual is the same, otherwise they may be the same or different.

Note: an individual is also called an instance. An individual is equivalent to an instance of a class.

3.2 Property)

An attribute is a binary relationship between individuals. That is, an attribute connects two individuals. For example, attributes hassibling may connect Matthew and Gemma, while attributes haschild will connect Peter and Matthew; attributes can have reverse attributes (inverse ), for example, the opposite attribute of hasowner is isownedby. The attribute can be limited to only one value, that is, the so-called function attribute. The attribute can also be transitive) or symmetric Ric ).

Note: The attributes mentioned here are the slots in proté Gé. In the description logic, they are role ), in object-oriented methods such as UML, they are relational, and in formal expressions such as Grail, they are called attributes ).

Class)

Classes in Owl represent individual sets. OWL uses a formal (mathematical) method to precisely describe the conditions required by members of the class. For example, all cats in the field belong to the cat class. Classes can form a hierarchy through inheritance relationships. Child classes are special circumstances in the parent class. For example, considering animal and CAT classes, cat can be a subclass of animal (that is, animal is the parent class of CAT). This indicates that all cats are animals and all cat members are members of the animal class, if you are a cat, you are also an animal. The cat class is included by the animal class, and so on. An important feature of OWL-DL is that the (inclusive) Relationship between parent class and Child class can be automatically calculated by the inference engine.

Note: The term concept is sometimes used to replace classes. In fact, classes are a concrete manifestation of concepts.