Eshelman, Larry and McDermott, J. 1986. Mole: A knowledge acquisition tool that uses its head. In Proceedings AAAI-86 Fifth National Conference on Artificial Intelligence, Philadelphia. AAAI. 950--955.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....referenced by the advice and the needs of any particular planning problem. The utility of explicit self models has long been recognized in the field of knowledge acquisition. Starting with Teiresias [ Davis, 1982 ] a number of knowledge acquisition systems (e.g. ask [ Gruber, 1989 ] and mole [ Eshelman et al. 1986 ] have employed explicit models of the problem solver under construction to simplify knowledge entry. Knowledge is typically added in response to the system s failing to handle some new example. Unfortunately, new knowledge must be entered in a form that is very close to fully operational. ....

Eshelman, Larry and McDermott, J. 1986. Mole: A knowledge acquisition tool that uses its head. In Proceedings AAAI-86 Fifth National Conference on Artificial Intelligence, Philadelphia. AAAI. 950--955.

....based on the structure of the domain and or of the task. KNACK [13] has knowledge about the design of electromechanical systems and uses such knowledge to make sense of the data it elicits from written reports and from interviews with the domain expert. A more interesting approach is found in MOLE [14], where a heuristic classification model [15] is used to elicit and interpret the data. The above approach is based on the idea that the architecture of an expert system should rely on a more principled framework than a mere collection of frames and rules. For instance, heuristic classification ....

....concerning top down approaches is that when conceptual models are used Motta p. 10 to drive the acquisition of knowledge, it may be the case that the model and the domain don t match. This is especially true for systems that implement automated knowledge acquisition (see for instance MOLE [14]) as they can only map the domain dependent data into the prestored model and provide little help when such a mapping is not feasible. To make things more complex, the expert may not be fully aware of his her assumptions and problem solving strategies. Therefore, it may be up to the knowledge ....

[Article contains additional citation context not shown here]

Eshelman, L., McDermott, J. MOLE: A Knowledge Acquisition Tool That Uses Its Head. Proceedings AAAI 86, Philadelphia, PA, 1986.

....technique emphasizes the construction of domain representations via the direct editing of knowledge in a semantic network, frame, or rule representation. It is one of the most popular paradigms, typified by tools as VITAL T211.4 OU 3, KSS0 Review, p. 5 The Open University 1991 diverse as MOLE (Eshelman et. al, 1986), KEATS (Motta et al. 1990, 1991) and KADS (Breuker Wielinga, 1985, 1989) The crucial insight of Gaines analysis is his characterization of precisely what is elicited from experts by this technique, namely: The distinctions that the expert makes about domain entities (attributes and ....

Eshelman, L. and McDermott, J. (1986). MOLE: A Knowledge Acquisition Tool that uses its Head. Proceedings of AAAI '86, Philadelphia, PA, 1986.

....model by means of repertory grids. In this case, we can say that the acquisition of knowledge is driven by a psychological model, to produce a conceptual representation of human expertise. SALT (Marcus and McDermott, 1989) which is based on a proposeand revise problem solving model, and MOLE (Eshelman and McDermott, 1986), which uses a heuristic classification model, also perform top down knowledge acquisition. However, as the models they employ are based on the architecture of the emerging expert system, we say that they are driven by design models, which are abstract descriptions of the behaviour of a ....

Eshelman, L., and McDermott, J. (1986). MOLE: A Knowledge Acquisition Tool that uses its Head. Proceedings of AAAI '86, Philadelphia, PA.

No context found.

Eshelman, L., McDermott, J. MOLE: A Knowledge Acquisition Tool That Uses Its Head. Proceedings AAAI 86, Philadelphia, PA, 1986.

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