Understanding Semantic Annotation – Part 3: Ontology and RDF

Tram Ho

To continue this series, today I will write about the two most basic concepts in Semantic Annotation: Ontology and RDF.


1.1 Concept

In computer science, an ontology is a data model that represents a field and is used to infer objects in that field and the relationship between them. Ontology provides a general vocabulary set of concepts, important attributes, and definitions of these concepts and attributes. In addition to the vocabulary set, ontologies provide constraints, sometimes referred to as basic assumptions about the meaning of the vocabulary set, which is used in a domain that can be communicated. between people and other heterogeneous distributed application systems. The ontology is used as a form to present knowledge about the world or part of it. The ontology often describes:

  • Individuals: Basic, foundation objects.
  • Classes: Collections, or types of objects.
  • Attributes: Attributes, features, characteristics, characteristics, or parameters that objects have and can be shared.
  • Relationships: The pathways by which objects can relate to another object. The ontology vocabulary set is built on the basis of the RDF and RDFS layers, providing flexible semantic representation for Web resources and the ability to support arguments.

1.2 Ontology components

Individuals – Instances Individuals are the basic, fundamental components of an ontology. Objects in an ontology can include specific objects such as people, animals, tables, etc. as well as abstract individuals such as members or words. An ontology may not need any individuals, but one of the main reasons for an ontology is to provide a semantics for classifying instances, even though they are not really part. of ontology. Classes – Concept Classes are groups of abstract objects. They may contain instances, other classes, or a combination of both. The ontologies vary depending on their structure and content: A class can contain subclasses, which can be an overview class (containing everything), maybe a class containing only individual items. , A class can be merged or merged by other classes. This lumped relationship is used to create a hierarchical structure of layers, usually with the most common Thing type at the top and the very explicit 2002 and Ford classes at the bottom. Properties Objects in ontology can be described by declaring their properties. Each attribute has its name and value. Attributes are used to store information that the object may have. For example, an individual may have attributes: FirstName, Date_date, Country_Que, Number_cmnd, etc. The value of an attribute can have complex data types. Relations One of the important applications of using attributes is to describe the relationships between objects in ontologies. A relationship is a valuable attribute that is an object in the ontology. An important type of relationship is the type of subsumption. This type of relationship describes which objects are members of which classes of objects. At the moment, combining ontologies is a process that is mostly done manually, thus very time consuming and expensive. The use of ontologies is fundamental to providing a common definition of core terms that can make this process more manageable. There are currently studies based on reproduction techniques to link ontologies, but this field is only theoretically available.

Example of 1 ontology

2- RDF

2.1 Concept

The content of the Web information is primarily for human consumption, and the machine cannot read and understand it. Therefore, it is difficult to automate any content on the Web, at least on a large scale. Moreover, with the vast amount of information on the Web, we cannot process them manually. Therefore, the W3C proposed a solution to describe data on the Web and can be understood by machines, which is RDF. RDF (short for Resource Description Framework, roughly translated as Resource Description Framework) originated in early 1999 by W3C organization as a standard for encoding metadata (metadata). The RDF name is officially introduced in W3C specification documents with a brief description.

2.2 Characteristics

RDF is a language for displaying information about web resources. (as per RDF Primer) RDF is a framework for displaying information on web (according to RDF Concept document) RDF is a general purpose language for displaying information on Web (according to RDF Syntax document and Data RDF Schema); RDF is a validation language used to describe prepositions using accurate official vocabulary, especially those that are specified using RDFS, to access and use the Web, and is intended to provide Level 1 basic platform for advanced validation languages ​​with the same purpose (according to RDF Semantics).

In a nutshell, Ontology is the basis of the vocabulary of semantics, and the language to describe ontologies can use RDF.

Below is an example of a book Ontology, described in RDF language

In the next section, I will talk more about graph data, which will make a difference in storage that makes the power of reasoning of semantic basis. ?

Share the news now

Source : Viblo