China
Africa
Explore chapters and articles related to this topic
A Problem Solving System for Data Analysis, Pattern Classification and Recognition
Published in Abraham Kandel, Gideon Langholz, Lotfi A. Zadeh, Hybrid Architectures for Intelligent Systems, 2020
Expert systems are often classified based on the type of knowledge representation used. There are two major knowledge-based expert systems: rule-based systems and frame-based systems. Rule-based systems use rules in the form of “IF conditions THEN actions”, the so-called production rule, to represent domain specific knowledge. During the execution of the rule-based system, a rule can fire, that is, can have its action part executed by the inference engine if its conditions are satisfied. Rule-based systems are the most frequently used architecture today because it is easy for production rules to represent empirical associations or rules of thumb explicitly. Also, by tracing the reasoning chain, simple explanation can be implemented conveniently. In general, it is also easy to add new production rules to the knowledge base. In frame-based systems, a typical situation or a class of objects are represented in a data structure called frame. This data structure includes both declarative information and procedural information in predefined internal relations. With either knowledge representation scheme it is possible to integrate abstract knowledge for decision making based on feature parameters extracted from images and invocation of procedures for low level image analysis required in inspection applications. Expert system shells provide sophisticated user interface tools in an environment that facilitate both system development and on-line usage.
Skill-Based Expert Systems in Robotics
Published in Spyros G. Tzafestas, Intelligent Robotic Systems, 2020
The skill-based expert system for robot control is a representative example of a nonnumerical modeling principle. It is therefore highly instructive to make a comparison between the two kind of models in terms of the preceding assessment criteria. As shown later, the skill-based expert system relies to a large extent on production rules. Accordingly, the basic set operation on which the control mechanism of skill control is built involves only the mapping relation S → P, where S is the set of inputs and P the set of outputs. As seen, restrictive assumptions for model building are of a most general nature. Consequently, the application field of such models is very wide and the extension potential to large systems practically unlimited if combined with multilevel control. Constraints of this type of nonnumerical models in the real-time control of large systems are also less strict than in the case of state space models. On the other side, the mapping relation cannot compare with vector spaces in the study of the dynamic properties of a plant.
Knowledge taxonomy vs. knowledge ontology and representation
Published in Jay Liebowitz, Knowledge Management: Learning from Knowledge Engineering, 2001
Another type of ontology is called a knowledge representation ontology. This is derived from the knowledge engineering field and can be in terms of mainly production rules, frames, semantic networks, and cases/scripts. Production rules are in the form of IF–THEN expressions, referred to as antecedents–consequents or conditions–actions. An example is: IF my nose is runningand I am sneezingand my throat is soreTHEN there is a good likelihood that I have a cold
Online emergency mapping based on disaster scenario and data integration
Published in International Journal of Image and Data Fusion, 2021
Fu Ren, Yiwen Li, Zhihe Zheng, Han Yan, Qingyun Du
The mapping rules are derived from the cartographic knowledge of the cartographer, including explicit conceptual knowledge and implicit experience knowledge (Ren et al. 2020). The design of disaster data mapping rules is closely related to various disaster scenarios. Thus, two aspects need to be considered. a) Disaster theory. Conducting corresponding evaluation and analysis based on research of disaster type, disaster state, disaster spatial scale, disaster event and emergency response scope, and manifestation of emergency response b) Cartographic knowledge: The theoretical and technical knowledge involved in conventional cartography and online cartography, which need to be acquired by learning relevant map standards and specifications, studying existing map works, and obtaining the experience and knowledge of cartographic experts. These rules are described in the computer in the form of production rules and stored as one of the knowledge bases. This paper designs the following three types of mapping knowledge rules for disaster data expression:
Function block-enabled operation planning and machine control in Cloud-DPP
Published in International Journal of Production Research, 2023
Mohammad Givehchi, Yongkui Liu, Xi Vincent Wang, Lihui Wang
Operation planning for MFs is a knowledge-intensive and complex decision-making process that is traditionally performed by experienced operation planners who are experts of the field. One source of knowledge in this field is scientific and experimental information provided in formulated and structured formats. They also might be converted to recommendations to be used in operation planning. Currently, there is no standard method for formulating and structuring knowledge for operation planning that applies to all cases, domains and users universally. To create an automatic component for operation planning, the formulation and structuring should be such that leads to support of dynamic, flexible knowledge-based planning rather than rigidity of the system. In fact, knowledge capturing from domain experts is impossible without some level of structuring (Park 2003). However, the system should have flexibility and supporting elements for a wider diversity of ways that information and logic are presented. Several research efforts have been made with respect to CAPP and computer-aided manufacturing (CAM) in the area of feature-based machining and knowledge-based operation planning (Refer to (Gupta and Ghosh 1989; Marri, Gunasekaran, and Grieve 1998; Xu, Wang, and Newman 2011; Kumar 2017, 2019)). An automatic operation planning component can use artificial intelligence methodologies to imitate human decision-making processes. Production rules systems are one of the disciplines in artificial intelligence that is able to capture and support explicit knowledge representation of domain experts in the form of rules and imitate the logical inference of humans based on the knowledge.