China
Africa
Explore chapters and articles related to this topic
Knowledge Representation and Decision Making for Mobile Robots
Published in Shuzhi Sam Ge, Frank L. Lewis, Autonomous Mobile Robots, 2018
Elena Messina, Stephen Balakirsky
A common type of symbolic representation for representing rules is ontological. Ontologies are definitions and organizations of classes of facts and formal rules for accessing and manipulating (and possibly extending) those facts. There are two main approaches to creating ontologies, one emphasizing the organizational framework, with data entered into that framework, and the other emphasizing large-scale data creation with relationships defined as needed to relate and use that data. Cyc [36] is an example of the latter, an effort to create a system capable of common sense, natural language understanding, and machine learning.
Knowledge Sharing and Reuse
Published in Jay Liebowitz, The Handbook of Applied Expert Systems, 2019
Although Cyc goals have remained unchanged over the years, Cyc technology has evolved substantially since then2. Actually, Cyc is a very large, multicontextual knowledge base with inference engines upon which different applications are built. Cyc technology consists of three main parts: the Cyc Knowledge Base (that is, the Cyc ontology), the CycL representation language and inference engines, and the knowledge server utility. The Cyc system is available in Common Lisp and C.
Knowledge in AI
Published in Janet Finlay, Alan Dix, An Introduction to Artificial Intelligence, 2020
The CYC project aims to build a knowledge base containing the millions of pieces of common knowledge that humans possess. It is a ten-year project involving many people, meticulously encoding the type of facts that are “obvious” to us, facts at the level of “all men are people” and “children are always younger than their parents”. To us, expressing such facts seems ludicrous; for the computer they need to be represented explicitly.
Structuring superconductor data with ontology: reproducing historical datasets as knowledge bases
Published in Science and Technology of Advanced Materials: Methods, 2023
Masashi Ishii, Koichi Sakamoto
Ontology research has been developed under several top ontologies, including the Basic Formal Ontology (BFO) [13], Cyc [14], and the Descriptive Ontology for Linguistic and Cognitive Engineering (DOLCE) [15]. Attempts to integrate general concepts and improve the knowability of local topics in the field of materials science include the Materials Ontology by Ashino [16], European Materials and Modelling Ontology (EMMO) [17], and MatOnto [18] based on DOLCE. While providing a global view of materials science is undoubtedly important, local ontologies based on specialized knowledge is important when considering solutions to specific issues close to practical use. However, it is essential to ensure logical consistency of the local ontology with the global one. In fact, there is no clear boundary between global and local, and a number of intermediates overlap to form concepts, such as ChEBI [19] (which deals with biologically interesting molecules in accordance with BFO), CHEMINF [20] (which provides an overview of chemical information), CHMO [21] (which formalizes experimental chemistry methods), and RXNO [22] (which deals broadly with chemistry). These intermediate ontologies deal with more general concepts than the ontologies addressed in this study. The ceramics ontology [23] and the crystal defect ontology [24] are examples of local ontologies that are as specialized as that in this study.
A ladder to human-comparable intelligence: an empirical metric
Published in Journal of Experimental & Theoretical Artificial Intelligence, 2018
The Cyc system (Lenat, Guha, Pittman, Pratt, & Shepherd, 1990), however, is based on different logic than the self-adaptive systems and is therefore ill-suited for LM measurement. Nonetheless, we can observe that it effectively passes the level (empty knowledge base at the start of the project!); its rules, however, are written by hand, or are self-induced by the inference engine, thus Cyc does not need to pass the higher levels sequentially to get to the TT itself eventually! Because the Cyc already delivers results, it is probably much further than ; thus, it may just be in front of current state-of-the-art ANN-based AGI designs. However, until any design passes the TT level, this is all a speculation.