Santos, Jr., Eugene. "A Fully Integrated Probabilistic Framework for Expert Systems Development. " Research proposal from Air Force Institute of Technology to Air Force Office of Scientific Research. March 1993.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....problems in design and implementation continue. Two such problems are the areas of uncertainty and knowledge acquisition (KA) Both seem inextricably linked; even if we could solve the problem of uncertainty, the problem of getting the desired knowledge into the system would be no easier to attain[21]. With disagreements among experts, as well as the existence of unknown knowledge 1 , it is possible that KA would be no easier to attain with consistency methods in place. Assuming, however, knowledge could be elicited, the knowledge itself must then be validated and verified to ensure the ....

Santos, Jr., Eugene. "A Fully Integrated Probabilistic Framework for Expert Systems Development. " Research proposal from Air Force Institute of Technology to Air Force Office of Scientific Research. March 1993.

.... (Generic Expert System Intelligent Assistant) is an intelligent user interface agent architecture conceived out of the development of a generic expert system shell (Harrington, Banks, Santos 1996) This expert system shell, called PESKI (Probabilities, Expert Systems, Knowledge, and Inference) (Santos 1993), is a collection of expert system tools under one architecture that is designed to be totally independent of any application domain. The tools contained in PESKI include an inference engine (IE) a knowledge base (called a Bayesian knowledge base (Banks 1995; Santos Santos 1996) or BKB) ....

Santos, Jr., E. 1993. A fully integrated probabilistic framework for expert system development. Technical report, Department of Electrical and Computer Engineering, Air Force Institute of Technology, WrightPatterson AFB, OH.

.... on one particular representation scheme known as Bayesian Knowledge Bases (BKBs) This BKB representation scheme is part of an overall expert system shell known as Probabilities, Expert System, Knowledge, and Inference 1 (PESKI) which is an integrated framework for expert system development [28]. This research focuses on developing a methodology to correct one problematic area of V V, namely unintentional incompleteness that may be present in the knowledge base. The results are currently integrated into the PESKI system. The incompleteness is recognized in the validation phase, and a ....

Santos, Jr., Eugene. "A Fully Integrated Probabilistic Framework for Expert Systems Development." Research proposal from Air Force Institute of Technology to Air Force Office of Scientific Research. March 1993.

....a backround of the system, the structure it uses to store knowledge, and the class of algorithms it employs for reasoning. 2. 1 PESKI AFIT is developing the Probabilistic, Expert System Knowledge and Inference (PESKI) environment, a domain independent architecture for building expert systems [19, 20, 21]. Figure 2.1 shows an overview of PESKI. The system is designed to store domain knowledge in the form of a probabilistic network model. It provides tools for knowledge acquisition and validation to improve the quality of the representation. PESKI also provides tools for reasoning over the stored ....

Santos, Jr., Eugene. A Fully Integrated Probabilistic Framework for Expert Systems Development . Technical Report, Department of Electrical and Computer Engineering, Air Force Institute of Technology, 1993.

....that, when combined under one system, provide the required functionality. All these tools share one thing in common: they all utilize and support a new knowledge representation called a Bayesian Knowledge Base (BKB) 75, 13] There are six basic tools that support PESKI s knowledge representation [18, 33, 74]. The knowledge acquisition (KA) tool is primarily responsible for putting knowledge into the BKB. The edit supports (ES) tool supplements the knowledge acquisition tool. The inference tool (IE) is responsible for extracting knowledge from the BKB. The verification and validation (VV) tool is ....

Santos, Jr., Eugene. A Fully Integrated Probabilistic Framework for Expert System Development . Technical Report, Department of Electrical and Computer Engineering, Air Force Institute of Technology, Wright-Patterson AFB, OH, 1993.

....: 6 1 Bibliography : BIB 1 Vita : VITA 1 iv List of Figures Figure Page 1.1. The four major subsystems of an expert system[26]. 1 2 1.2. The peski architecture[26] 1 5 2.1. A simple Bayesian Network : 2 2 2.2. A simple Bayesian Knowledge Base : 2 3 3.1. A BKB Polytree : ....

....: BIB 1 Vita : VITA 1 iv List of Figures Figure Page 1.1. The four major subsystems of an expert system[26] 1 2 1.2. The peski architecture[26]. 1 5 2.1. A simple Bayesian Network : 2 2 2.2. A simple Bayesian Knowledge Base : 2 3 3.1. A BKB Polytree : 3 2 3.2. The General BKB ....

[Article contains additional citation context not shown here]

Santos, Jr., Eugene. "A Fully Integrated Probabilistic Framework for Expert Systems Development." Research proposal from Air Force Institute of Technology to Air Force Office of Scientific Research Grant #94-0006, 22 October 1993.

....networks in intelligent user interfaces. GESIA (Generic Expert System Intelligent Assistant) 6] is an intelligent user interface architecture conceived out of the development of a generic expert system. This expert system, called PESKI (Probabilities, Expert Systems, Knowledge, and Inference) [11], is a collection of expert system tools under one architecture that is designed to be totally independent of any application domain, i.e. generic. The tools contained in PESKI include an inference engine, a knowledge base (called a Bayesian knowledge base [1] or BKB) a knowledge acquisition ....

Eugene Santos, Jr. A fully integrated probabilistic framework for expert system development. Technical report, Department of Electrical and Computer Engineering, Air Force Institute of Technology, Wright-Patterson AFB, OH, 1993.

....of each algorithm. In this paper we present such an architecture for a fast inference engine. This architecture, called overmind can be readily implemented using existing packages such as pvm [14] and genesis [16] The reasoning engine discussed here, as part of a larger system called peski [27] is currently being used for engine diagnostics for the Space Shuttle Program [2] However, for the purposes of this paper, we concentrate mainly on the reasoning architecture. The rest of this paper provides some background on probabilistic reasoning, presents the overmind architecture, ....

Eugene Santos, Jr. A fully integrated probabilistic framework for expert systems development. AFOSR Grant #940006, 1994.

No context found.

Santos, Jr., Eugene. "A Fully Integrated Probabilistic Framework for Expert Systems Development. " Research proposal from Air Force Institute of Technology to Air Force Office of Scientific Research. March 1993.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC