Knowledge representation and reasoning

Knowledge representation (KR) aims to model information in a structured manner to formally represent it as knowledge in knowledge-based systems whereas knowledge representation and reasoning (KRR, KR&R, or KR²) also aims to understand, reason, and interpret knowledge. KRR is widely used in the field of <a href="/facts/Artificial_intelligence/lJGauQwX">artificial intelligence</a> (AI) with the goal to represent <a href="/facts/Information/3LdrzXtJ">information</a> about the world in a form that a computer system can use to solve complex tasks, such as <a href="/facts/Computer-aided_diagnosis/M88tS5ax">diagnosing a medical condition</a> or <a href="/facts/Natural_language_user_interface/TSsXHy2V">having a natural-language dialog</a>. KR incorporates findings from psychology about how humans solve problems and represent knowledge, in order to design <a href="/facts/Formal_system/KaeDAjK7">formalisms</a> that make complex systems easier to design and build. KRR also incorporates findings from <a href="/facts/Logic/8JnpJLtt">logic</a> to automate various kinds of reasoning.
Traditional KRR focuses more on the declarative representation of knowledge. Related knowledge representation formalisms mainly include <a href="/facts/Vocabulary/gUfoY6TJ">vocabularies</a>, <a href="/facts/Thesaurus/w7xnH66H">thesaurus</a>, <a href="/facts/Semantic_network/mt7FlP53">semantic networks</a>, <a href="/facts/Axiom_system/dwMz70d5">axiom systems</a>, <a href="/facts/Frame_(artificial_intelligence)/YnQaR34X">frames</a>, <a href="/facts/Rule-based_system/xmlMNr7j">rules</a>, <a href="/facts/Logic_programming/Xgh8lj3F">logic programs</a>, and <a href="/facts/Ontology_(information_science)/SQlhl6lp">ontologies</a>. Examples of <a href="/facts/Automated_reasoning/x5EdkKoU">automated reasoning</a> engines include <a href="/facts/Inference_engine/oLlNX14Q">inference engines</a>, <a href="/facts/Automated_theorem_proving/vKGvuKFI">theorem provers</a>, <a href="/facts/Boolean_satisfiability_problem/wBVRjWoy">model generators</a>, and <a href="/facts/Statistical_classification/jXXHRkXR">classifiers</a>.
In a broader sense, parameterized models in <a href="/facts/Machine_learning/e0w0XJTu">machine learning</a> — including <a href="/facts/Neural_network_(machine_learning)/6V1jMlkx">neural network</a> architectures such as <a href="/facts/Convolutional_neural_network/kHEdnmGU">convolutional neural networks</a> and <a href="/facts/Transformer_(deep_learning_architecture)/cDbjx6a8">transformers</a> — can also be regarded as a family of knowledge representation formalisms. The question of which formalism is most appropriate for knowledge-based systems has long been a subject of extensive debate. For instance, Frank van Harmelen et al. discussed the suitability of logic as a knowledge representation formalism and reviewed arguments presented by anti-logicists. Paul Smolensky criticized the limitations of symbolic formalisms and explored the possibilities of integrating it with connectionist approaches.
More recently, Heng Zhang et al. have demonstrated that all universal (or equally expressive and natural) knowledge representation formalisms are recursively isomorphic. This finding indicates a theoretical equivalence among mainstream knowledge representation formalisms with respect to their capacity for supporting <a href="/facts/Artificial_general_intelligence/vbxwaq8t">artificial general intelligence</a> (AGI). They further argue that while diverse technical approaches may draw insights from one another via recursive isomorphisms, the fundamental challenges remain inherently shared.

Knowledge representation and reasoning open-in-new

Knowledge representation and reasoning