Knowledge representation and knowledge engineering allow AI programs to answer questions intelligently and make deductions about real world facts. Knowledge representation is a very important concept in expert systems and artificial intelligence (AI) in general. It involves the consideration of intelligent (expert) systems and how it presents knowledge. It is best understood in term of the roles it plays based on the task at hand.
A representation of “what exists” is an ontology: the set of objects, relations, concepts, and properties formally described so that software agents can interpret them. The most general ontologies are called upper ontologies, which attempt to provide a foundation for all other knowledge and act as mediators between domain ontologies that cover specific knowledge about a particular knowledge domain (field of interest or area of concern). A truly intelligent program would also need access to commonsense knowledge; the set of facts that an average person knows. The semantics of an ontology is typically represented in a description logic, such as the Web Ontology Language.
AI research has developed tools to represent specific domains, such as: objects, properties, categories and relations between objects; situations, events, states and time; causes and effects; knowledge about knowledge (what we know about what other people know); default reasoning (things that humans assume are true until they are told differently and will remain true even when other facts are changing); as well as other domains. Among the most difficult problems in AI are: the breadth of commonsense knowledge (the number of atomic facts that the average person knows is enormous); and the sub-symbolic form of most commonsense knowledge (much of what people know is not represented as “facts” or “statements” that they could express verbally).
Formal knowledge representations are used in content-based indexing and retrieval, scene interpretation, clinical decision support, knowledge discovery (mining “interesting” and actionable inferences from large databases), and other areas.
In artificial intelligence, knowledge representation is the study of how the beliefs, intentions, and value judgments of an intelligent agent can be expressed in a transparent, symbolic notation suitable for automated reasoning. From a purely computational point of view, the major objectives to be achieved are breadth of scope, expressivity, precision, support of efficient inference, learnability, robustness, and ease of construction. Knowledge-based techniques have been applied successfully for many computational tasks including text interpretation and cognitive robotics.
Many different general architectures have been used for knowledge representation, including first-order logic, other formal logics, semantic networks, and frame-based systems. The representation of temporal knowledge is both a problem of central importance in knowledge representation and an archetype of the kinds of issues that arise in developing representations for various domains. The use of machine learning techniques for the automatic construction of knowledge bases and knowledge representations is difficult, but has achieved some degree of success.
Not only are appropriate knowledge representations critical to the design and performance of commercially valuable software programs, our choice of knowledge representation systems also surfaces our (often) implicit theories about the very nature of machine and human intelligence. For instance, the proponents of rule-based representations and those of semantic nets may hold somewhat different views of how knowledge is organized in human minds.
While some AI practitioners restrict their attention to the development of software systems that perform particular tasks, the broader field of cognitive science asks questions about what constitutes knowledge and how it is organized in humans and machines. For such inquiry, knowledge representation systems provide a vehicle for expressing and testing theoretical claims and hypotheses.
Knowledge representation reasoning plays a central role in Artificial Intelligence. Research in Artificial Intelligence started off by trying to identify the general mechanisms responsible for intelligent behaviour. However, it quickly became obvious that general and powerful methods are not enough to get the desired result, namely, intelligent behavior. Almost all tasks a human can perform which are considered to require intelligence are also based on a huge amount of knowledge. For instance, understanding and producing natural language heavily relies on knowledge about the language, about the structure of the world, about social relationships, and so on.
One way to address the problem of representing knowledge and reasoning about it is to use some form of logic. While this seems to be a natural choice, it took a while before this ‘logical point of view’ became the prevalent approach in the area of knowledge representation. The important point about using formal logic is the logical method.
Faceted knowledge representation provides a formal mechanism for the implementation of knowledge systems. Faceted knowledge representation is also referred to as “basic unit,” “inter-relations,” “facet,” and “final interpretation.” Basically, facets occur as relational structures that combine units, and relations in which each facet stands for an aspect of a knowledge system. When these facets or relationships are interpreted, the resultant mappings can be used for translation/cross-mapping between different representations.
Faceted knowledge representation originates from the vision of designing a knowledge representation system that is applicable to a variety of domains and suits a variety of users. Here the system will provide a flexible means of coding and displaying knowledge structures depending on adjustable internal or user-defined facets. In the context of conventional knowledge representation, these appear close to formalisms, such as object-oriented design, DL, relational databases, formal concept analysis, and conceptual graphs.
The architecture of a faceted knowledge representation is based on a defined set of primitive notions, such as unit, relation, and facet, and an open set of logical and relational operators. Another important aspect is the combined extensional, set-oriented, and intentional, relation-oriented approach.
Knowledge representation involves representing the key concepts and relations between the decision variables in some formal manner, typically within a framework suggested by an expert systems shell. Most representation mechanisms must provide support for three aspects of knowledge, conceptual representation, relational representation, and uncertainty representation. As such, four schemes are commonly used for knowledge representation.
In general, the representation technique selected must be simple and intuitive to the task domain. The representation scheme selected must support full disclosure. In other words, the knowledge coded into the expert system must be simple to understand when examined by a person unfamiliar with the task domain.