Upon completion of this lesson, you will be able to:
An ontology is a representation of knowledge in a specific domain, that defines the concepts, objects, and relationships in that domain. The need for an ontology arises because it provides a shared and common understanding of the vocabulary, structure, and meaning of data within a domain, which is crucial for facilitating effective communication, data integration, and knowledge management. Ontologies play a crucial role in several fields such as artificial intelligence, information retrieval, semantic web, and many more, as they help in establishing a common semantic framework for the exchange and reuse of data and knowledge.
An ontology is a representation of a shared understanding of a domain of knowledge that defines the concepts and categories, and the relationships between them, in a clear and formally structured manner.
The need for an ontology arises in order to:
Facilitate communication: Ontologies provide a common vocabulary and set of concepts that allow individuals to communicate and exchange information more effectively.
Enhance interoperability: By providing a shared understanding of a domain, ontologies allow different systems and applications to work together seamlessly, improving interoperability and reducing redundancy.
Improve decision-making: Ontologies can be used to structure and organize information in a way that makes it easier to analyze and use for decision-making purposes.
Enable knowledge representation: Ontologies can be used to represent knowledge in a formal and structured manner, allowing for computer processing and manipulation of that knowledge.
Facilitate reuse and sharing of knowledge: Ontologies provide a common representation of knowledge that can be reused and shared across multiple applications and domains.
Before venturing further into the world of ontologies and ontology design, consider watching the lecture (40 min) by Dr. Martin Schedlbauer of Khoury Boston:
Concepts are fundamental to the design of an ontology, but are not simple to define – there is an absence of a uniform and generally accepted definition. In the context of ontology design, a “concept” refers to an abstract or general idea that represents a category or class of entities within a specific domain of knowledge. Concepts are fundamental building blocks in ontologies and are used to model and classify different types of objects, phenomena, or ideas within that domain.
The word concept comes from the Latin conceptum which means “that which is to be conceived”. So, concepts are as much ideas about how the mind organizes and perceives reality as it is a description of a physical or abstract object.
Key characteristics of a concept in ontology design include:
Abstraction: Concepts are often abstract and represent a higher-level notion that can have multiple instances or members. They capture the essential characteristics shared by all instances falling under the same category.
Categorization: Concepts are used to categorize and group similar entities or instances together based on their shared attributes, properties, or relationships.
Hierarchy: Concepts can be organized hierarchically, forming a taxonomy or ontology structure. This hierarchy represents broader, more general concepts at higher levels and more specific, narrower concepts at lower levels.
Properties and Relationships: Concepts can have associated properties and relationships that describe their attributes, characteristics, and how they relate to other concepts within the ontology.
For example, in a medical ontology, “Patient” could be a concept that represents individuals seeking medical care. This concept may have properties such as “Name,” “Date of Birth,” and “Medical Record Number.” Additionally, it may be related to other concepts like “Diagnosis” or “Treatment” through various relationships.
Concepts serve as the foundation for creating a structured and formal representation of knowledge within an ontology. They enable the organization, classification, and modeling of domain-specific information, making it easier for both humans and machines to understand, query, and reason about the domain’s knowledge.
From the perspective of computational information science, a concept is represented by a projection of a concept’s relevant attributes onto a digital representation as either a class definition in an object-oriented language, an externalization in XML, or a visual representation in UML.
Terms and facts are foundational to the creation of an ontology for a domain.
In the context of an ontology, a term is the word used to refer to the concepts represented within the ontology. These terms are used to describe and categorize the objects or phenomena in the domain of interest.
Ontologies are described using language. A term is a word drawn from that language that is agreed upon to represent that concept. If no word has yet been ascribed to a concept, then either a new word is invented or a word is borrowed from another language. For example, the term for a personal wireless communication device using cellular technology is “Cell Phone” in US English, but in German it is referred to by the (made-up) term “Handy”. In British English as well as a synonym in US English, the term “Mobile Phone” is also used. So, a concept can be described by more than one word.
It can also be possible that a term describes multiple unrelated concepts. Consider the term “bank” (English language), which can describe two different concepts depending on the context:
Financial Institution Concept: In one context, “bank” refers to a financial institution where individuals and businesses can deposit money, obtain loans, manage their accounts, and conduct various financial transactions. This concept represents a place where people store and manage their money, and it’s commonly associated with services like checking and savings accounts, mortgages, and loans.
Natural Landform Concept: In another context, “bank” can refer to a natural landform along the side of a river. In this concept, a “river bank” represents the sloped or elevated area adjacent to a river or other bodies of water. River banks are often composed of soil, sand, or other materials and play a role in controlling water flow, erosion, and providing habitats for various organisms.
As we can see, the term “bank” can have different meanings and represent distinct concepts depending on whether it is used in a financial or environmental context.
Another example of the same word, but coming from different languages, can describe a different concept. For example, the term “old timer” describes an old person in English but an antique car in German.
Some languages require an adjective plus a noun rather than a single word to describe a single concept. For example, the word “madrugada” (Portuguese) describes the early morning hours from about 1am to 6am. In English, there is no single word, but rather English uses the expression “early morning hours”. On the other hand, German does have a single word to describe this period of the night: “Dämmerung”.
A common strategy is to define every term with a glossary definition and provide example instances and collect the terms and their different descriptive words in one or more languages along with synonyms in a catalog.
This illustrates how context and language are essential in understanding the intended meaning of a term, and why ontologies and knowledge representation systems aim to disambiguate such terms to ensure clarity and precision in communication and data interpretation.
The Sapir-Whorf hypothesis, also known as linguistic relativity or linguistic determinism, suggests that the language we use can influence and shape the way we perceive and think about the world. In the context of defining concepts as ideas shaped by language, the Sapir-Whorf hypothesis highlights the profound role that language plays in the formation and representation of our concepts and ideas.
Here’s how the Sapir-Whorf hypothesis relates to the definition of concepts:
Language and Conceptualization: According to the Sapir-Whorf hypothesis, the structure and vocabulary of a language can impact how individuals conceptualize and categorize the world around them. In other words, the words and grammatical structures available in a language can shape the way people think about and classify their experiences and observations.
Language as a Lens: Language serves as a lens through which we interpret and understand our environment. Different languages may have different words or linguistic constructs to describe and categorize phenomena, leading to variations in how concepts are formed and represented.
Cultural Variation: The Sapir-Whorf hypothesis suggests that linguistic differences can lead to cultural variations in how people perceive and categorize concepts. For example, some cultures may have more specific words or concepts for certain ideas or phenomena that are more generalized in other languages.
Implications for Ontology: When designing ontologies or knowledge representation systems, it’s important to consider the linguistic and cultural aspects of concepts. Concepts within an ontology are often labeled using language, and the choice of terminology can impact how users understand and interact with the ontology.
Translation Challenges: The hypothesis also raises issues related to translation between languages. Translating concepts and ideas from one language to another can be challenging because the nuances, connotations, and cultural associations of words may not align perfectly.
In short, the Sapir-Whorf hypothesis underscores the idea that our concepts and ideas are not universally fixed but are, to a significant extent, influenced by the language we use. This perspective has implications for how we approach knowledge representation and ontology design, as it highlights the need to consider linguistic and cultural factors when modeling and representing concepts within ontological frameworks.
Facts are statements or relationships that hold true between the terms in the ontology. For example, a fact in an ontology about animals could be that “a cat is a type of mammal.” These facts represent knowledge and provide information about the relationships between the terms in the ontology. In an ontology, the terms and facts are defined in a formal and structured way, making it possible for computer programs to process and use the information in a consistent and meaningful manner.
In an ontology, terms and facts are organized into a hierarchical structure and connected by relationships. This structure provides a way to represent the complex relationships and interconnections between concepts in a domain, making it easier to understand and use the knowledge represented in the ontology.
Terms and facts are often defined in a language but as language is fraught with ambiguity, structured narratives such as Structured English can be helpful in reducing ambiguity and forcing definitive discipline. See Lesson 50.204 – Ontology Description with Structured Narratives for a deeper look at how to write structured narratives.
In the context of ontology design, “relationships” refer to the associations, connections, or links that exist between different concepts or classes within the ontology. These relationships are used to represent how concepts are related to each other and to model the interactions and dependencies that exist within a specific domain of knowledge. Relationships play a crucial role in defining the structure and semantics of an ontology. They are a type of fact.
Here are some key aspects of relationships in ontology design:
Connecting Concepts: Relationships establish connections between concepts or classes in the ontology, allowing you to express how different entities or ideas are related to one another.
Types of Relationships: Relationships can take various forms and represent different types of associations. Common types of relationships include “is-a” (subsumption), “part-of,” “has-property,” “related-to,” and many others. These relationships help define the semantics of how concepts are interrelated.
Directionality: Relationships may have directionality. For example, in a “has-part” relationship, one concept (e.g., “Car”) may have parts (e.g., “Engine,” “Wheels”), but those parts are not themselves “Cars.” Understanding the direction of relationships is essential for accurate modeling.
Cardinality: Relationships can specify the number of instances involved in the association. For instance, a “has-child” relationship between a “Family” concept and a “Child” concept may indicate that a family can have multiple children, implying a cardinality of “one-to-many.”
Inverse Relationships: Many ontologies include inverse relationships to represent the reverse direction of a connection. For example, if there is a “parent-of” relationship, there can also be an “child-of” inverse relationship.
Transitivity and Symmetry: Some relationships may exhibit transitivity (if A is related to B and B is related to C, then A is related to C) or symmetry (if A is related to B, then B is related to A). These properties can be important for reasoning and inference.
Property Characteristics: Relationships can have associated properties or attributes that further describe the nature of the relationship. These properties can include metadata about the relationship, such as its date of creation, creator, or other relevant information.
In ontology design, defining relationship facts between concepts is essential for accurately representing the knowledge and domain-specific semantics. Properly modeling these relationships helps facilitate reasoning, querying, and data integration within the ontology, making it a valuable tool for knowledge organization and retrieval in various applications, including information retrieval, data integration, and artificial intelligence.
There are two relationships that are special in that they define hierarchy:
A partonomy, also known as a part-whole or whole-part hierarchy, is a type of relationship used in ontology design to represent the relationship between a whole and its parts. In a partonomy, a whole is defined as an object that is composed of other objects, which are referred to as parts.
For example, in an ontology of objects, a “car” might be represented as a whole and its various parts, such as the engine, wheels, and body, would be represented as parts. The partonomy relationship is a crucial aspect of ontology design as it provides a way to represent the composition and structure of objects and concepts in a domain.
Partonomies can be used to support a variety of reasoning tasks, such as classification, object recognition, and object manipulation, as well as to represent the hierarchical structure of concepts in a domain. They are commonly used in fields such as artificial intelligence, knowledge representation, and semantic web.
Taxonomy, in the context of an ontology, refers to the process of defining a more general concept by combining multiple more specific concepts. This relationship between concepts is also known as the is-a relationship, a subtype-supertype relationship, or a generalization hierarchy.
For example, in an ontology of animals, the concept of “mammal” could be defined as a generalization of more specific concepts, such as “dog,” “cat,” and “horse.” The generalization relationship captures the idea that dogs, cats, and horses are all mammals, and thus have certain properties and characteristics in common.
Taxonomies are used in ontology design to organize concepts into a hierarchical structure and to capture the relationships between concepts in a domain. This structure provides a way to represent the complex relationships and interconnections between concepts and supports reasoning tasks such as classification and inference.
The use of taxonomies in ontologies also makes it possible to reuse existing knowledge and to define new knowledge in a more concise and manageable way. This can be useful in a variety of domains, including artificial intelligence, knowledge representation, and semantic web.
Taxonomy as a generalization relationship is closely related to inheritance in object-oriented design. Inheritance is the programming language’s way of representating of a taxonomy.
A visual representation of an ontology is a graphical representation of the concepts and relationships in an ontology. It is used to provide a visual representation of the structure of the knowledge represented in the ontology and to facilitate the understanding and interpretation of the ontology.
There are several ways to visually represent an ontology, including:
Concept maps: A concept map is a visual representation of the concepts and relationships in an ontology, showing the hierarchy and relationships between concepts.
Entity-relationship diagrams (ERDs): ERDs are diagrams that represent the concepts in an ontology as entities and the relationships between them as lines connecting the entities.
UML Class diagrams: Class diagrams are used in object-oriented programming and also in ontology design. They represent the classes (or concepts) in the ontology and the relationships between them.
Tree diagrams: Tree diagrams are hierarchical diagrams that show the relationships between concepts in an ontology as a tree structure.
Network diagrams: Network diagrams represent the concepts and relationships in an ontology as nodes (representing concepts) and edges (representing relationships).
The choice of visual representation depends on the particular goals and needs of the ontology and the audience for the representation. Some visual representations may be more suited to specific tasks, such as reasoning or navigation, while others may be more suited to presenting the overall structure of the ontology. Regardless of the visual representation used, the goal is to provide a clear, concise, and meaningful representation of the knowledge in the ontology. It is not unusual to use more than one visual representation and to present different visual models to different stakeholders.
A concept map in ontology design is a graphical representation of the concepts and relationships in an ontology. It is used to provide a visual representation of the structure of the knowledge represented in the ontology and to facilitate the understanding and interpretation of the ontology.
A concept map typically consists of nodes, which represent concepts, and lines connecting the nodes, which represent relationships between the concepts. The nodes are typically organized in a hierarchical structure, showing the relationships between more general and more specific concepts. The lines connecting the nodes can represent different types of relationships, such as “is-a” relationships, “part-of” relationships, or other relationships specific to the domain represented in the ontology.
Concept maps provide a visual representation of the structure of the knowledge in an ontology, making it easier to understand and use the knowledge represented in the ontology. They are often used in education and training to help learners understand complex concepts and relationships, and they can be useful in a variety of domains, including artificial intelligence, knowledge representation, and semantic web.
Overall, concept maps are a powerful tool in ontology design for representing knowledge in a clear, concise, and meaningful way, and for facilitating the understanding and use of the knowledge represented in the ontology.
In the context of visual ontology modeling, Entity-Relationship Diagrams (ERDs) are graphical representations used to depict the structure and relationships of entities (or concepts) and their associations within an ontology or a database. ERDs are commonly used in database design and ontology modeling to illustrate the organization and connectivity of data or knowledge elements.
Key components and concepts related to ERDs include:
Entities: In ontology modeling, entities represent concepts, classes, or categories of objects within a specific domain. Each entity is typically depicted as a rectangle in the ERD.
Attributes: Attributes represent properties or characteristics of entities, providing additional information about them. Attributes are usually displayed within the entity rectangle.
Relationships: Relationships describe how entities are connected or associated with one another. Lines or connectors between entities in the ERD illustrate these relationships and their cardinality (e.g., one-to-one, one-to-many, many-to-many).
Keys: Keys are used to uniquely identify instances of an entity. In ERDs, primary keys are often designated to ensure the uniqueness of each entity instance.
Associations: Associations represent connections between entities without specifying a direction or specific cardinality, often used in ontologies to capture general relationships.
ERDs are valuable in ontology modeling for several reasons:
Visualization: ERDs provide a visual representation of the ontology’s structure, making it easier for stakeholders to understand the relationships and dependencies among entities and concepts.
Communication: ERDs serve as a communication tool between domain experts, ontology designers, and other stakeholders, facilitating discussions about the structure and semantics of the ontology.
Design and Validation: ERDs help in the design and validation of ontologies by visualizing potential issues, such as redundant relationships or missing entities.
Documentation: ERDs can be used as documentation for the ontology, offering a concise and comprehensible summary of its structure.
Implementation: ERDs can serve as a blueprint for implementing the ontology in a knowledge base or database system.
Overall, Entity-Relationship Diagrams are a practical visual tool for modeling and representing the structure and relationships within ontologies, aiding in the design, development, and maintenance of knowledge systems and databases.
Unified Modeling Language (UML) Class Diagrams are another visual modeling tool that can be used in the visualization of ontology designs and knowledge representations, particularly when modeling object-oriented ontologies or systems. UML Class Diagrams are primarily associated with software engineering and object-oriented design, but they can also be adapted for ontology modeling.
Classes: In UML Class Diagrams, classes represent concepts, entities, or types of objects in the system or ontology. These classes can be equated with ontological concepts or entities.
Attributes: Attributes within UML classes correspond to the properties or characteristics of ontological concepts. They describe the data or properties associated with the classes.
Associations: Associations in UML Class Diagrams depict relationships between classes, similar to how relationships represent connections between concepts in ontologies. Associations can have multiplicities (e.g., one-to-one, one-to-many) to capture cardinality.
Inheritance: UML Class Diagrams support the modeling of inheritance relationships, where one class (subclass) inherits attributes and behaviors from another class (superclass). In ontology design, this can represent concepts and subconcepts.
Generalization/Specialization: Generalization relationships in UML illustrate how one class (the specialized or subclass) is a more specific version of another class (the generalized or superclass), which aligns with hierarchical taxonomies in ontologies.
Multiplicity: Multiplicity in UML Class Diagrams specifies how many instances of a class can be associated with instances of another class, mirroring cardinality constraints in ontologies.
Aggregation and Composition: UML Class Diagrams can represent aggregation and composition relationships to show how parts are related to a whole. This can be relevant when modeling part-whole relationships in ontologies.
While UML Class Diagrams are originally designed for software modeling, they offer a way to visually represent ontologies with an emphasis on object-oriented concepts and relationships. However, for complex ontologies and semantic web applications, specialized ontology modeling languages like OWL (Web Ontology Language) are often preferred due to their semantic expressiveness and reasoning capabilities. Nevertheless, UML Class Diagrams can be a useful tool for capturing high-level ontological structures and relationships, especially when transitioning from software design to ontology development.
Tree diagrams in the context of ontology design are a type of visual representation of an ontology. They are hierarchical diagrams that show the relationships between concepts in an ontology as a tree structure.
In a tree diagram, the concepts in an ontology are represented as nodes, and the relationships between concepts are represented as lines connecting the nodes. The nodes are organized in a hierarchical structure, with more general concepts at the top of the tree and more specific concepts at the bottom of the tree. The lines connecting the nodes represent “is-a” relationships, where a more specific concept is a subtype of a more general concept.
Tree diagrams are often used to represent the hierarchical structure of the knowledge in an ontology, and to provide a clear, concise, and meaningful representation of the relationships between concepts in the ontology. They are useful for visualizing the relationships between concepts and for exploring the structure of the knowledge in an ontology.
Overall, tree diagrams are a powerful tool in ontology design for representing the hierarchical structure of knowledge in a clear, concise, and meaningful way, and for facilitating the understanding and use of the knowledge represented in the ontology.
Network diagrams in the context of ontology design are a type of visual representation of an ontology that shows the relationships between concepts as a network of interconnected nodes.
In a network diagram, the concepts in an ontology are represented as nodes, and the relationships between concepts are represented as lines connecting the nodes. The nodes can represent different types of relationships, such as “is-a” relationships, “part-of” relationships, or other relationships specific to the domain represented in the ontology. The lines connecting the nodes can also be labeled to indicate the type of relationship being represented.
Network diagrams provide a more flexible representation of the relationships between concepts in an ontology, as they can show multiple relationships between concepts, as well as more complex relationships between concepts. They are useful for visualizing the relationships between concepts in an ontology, and for exploring the structure of the knowledge represented in the ontology.
Overall, network diagrams are a powerful tool in ontology design for representing the relationships between concepts in an ontology in a clear, concise, and meaningful way, and for facilitating the understanding and use of the knowledge represented in the ontology.
Ontology design is the process of creating a formal representation of a set of concepts and their relationships within a specific domain. An ontology is a structured representation of knowledge that captures the concepts and relationships within a domain, and provides a way to represent the structure of the knowledge in a clear, concise, and meaningful way.
The goal of ontology design is to create a comprehensive and accurate representation of the knowledge in a domain, and to make this knowledge accessible and usable for a variety of purposes, such as reasoning, navigation, and information retrieval.
The process of ontology design involves identifying the concepts and relationships in a domain, defining the concepts and relationships in a formal representation, and organizing the concepts and relationships into a hierarchical structure. The representation of concepts and relationships in an ontology can include terms, definitions, axioms, and other elements that provide a precise and unambiguous representation of the knowledge in the domain.
Ontology design is used in a variety of domains, including artificial intelligence, knowledge representation, semantic web, and others, and can be represented using a variety of visual representations, such as concept maps, entity-relationship diagrams, class diagrams, tree diagrams, and network diagrams.
Overall, ontology design is a crucial step in the process of creating knowledge-based systems, and is important for creating a clear, concise, and meaningful representation of the knowledge in a domain, and for making this knowledge accessible and usable for a variety of purposes.
Jakus, G., Milutinović, V., Omerović, S., Tomažič, S., Jakus, G., Milutinović, V., & Tomažič, S. (2013). Concepts, Ontologies, and Knowledge Representation. Springer New York.