Eliciting knowledge and transferring it effectively to a knowledge-based system

Gaines, Brian R.; Shaw, Maria C.

doi:10.1109/69.204087

Cited by 102 publications

(44 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This knowledge must be acquired: either deduced from physical laws, from designers in person or induced from analysis of previous designs. This knowledge acquisition is a known bottleneck for developing expert systems (Gaines and Shaw, 1993), and by extension design support tools. Methodologies have been developed to acquire design knowledge from domain experts, but this tends to be highly tailored to the particular domain under review (Hughes et al, 2001).…”

Section: Knowledge Acquisitionmentioning

confidence: 99%

Challenges to Bayesian decision support using morphological matrices for design: empirical evidence

Matthews

2010

Res Eng Design

View full text Add to dashboard Cite

Section: Knowledge Acquisitionmentioning

confidence: 99%

Challenges to Bayesian decision support using morphological matrices for design: empirical evidence

Matthews

2010

Res Eng Design

View full text Add to dashboard Cite

“…However they are also time-consuming tasks. AI researchers who have been interested in eliciting qualitative knowledge detained by the experts, know the difficulties of this task [11]. Therefore expert intervention must remain limited, and be guided to be efficient.…”

Section: Introductionmentioning

confidence: 99%

An iterative approach to build relevant ontology-aware data-driven models

Thomopoulos

Destercke

Charnomordic

et al. 2013

Information Sciences

View full text Add to dashboard Cite

In many fields involving complex environments or living organisms, data-driven models are useful to make simulations in order to extrapolate costly experiments and to design decision-support tools. Learning methods can be used to build interpretable models from data. However, to be really useful, such models must be trusted by their users. From this perspective, the domain expert knowledge can be collected and modelled to help guiding the learning process and to increase the confidence in the resulting models, as well as their relevance. Another issue is to design relevant ontologies to formalize complex knowledge. Interpretable predictive models can help in this matter. In this paper, we propose a generic iterative approach to design ontology-aware and relevant data-driven models. It is based upon an ontology to model the domain knowledge and a learning method to build the interpretable models (decision trees in this paper). Subjective and objective evaluations are both involved in the process. A case study in the domain of Food Industry demonstrates the interest of this approach.

show abstract

“…Our early studies in the 1980s targeted expert system shells as inference engines (Gaines and Linster, 1990) and found that many extraneous structures had to be added to manage what was not logically well-founded inference (Gaines and Shaw, 1993). With the development of logical foundations for KL-ONE semantic network systems (Brachman, 1977), leading through various implementations to the well-defined CLASSIC specification (Borgida, et al, 1989), it became attractive to target logically sound knowledge representation systems, and we implemented a formal visual language (Gaines, 1991b) for the semantics of CLASSIC, and a knowledge representation server, KRS, that was a modular, extensible, C++ implementation of CLASSIC based on an algebraic model of the logical semantics (Gaines, 1993).…”

Section: Introductionmentioning

confidence: 99%

Designing Visual Languages for Description Logics

Gaines

2008

J of Log Lang and Inf

View full text Add to dashboard Cite

Semantic networks were developed in cognitive science and artificial intelligence studies as graphical knowledge representation and inference tools emulating human thought processes. Formal analysis of the representation and inference capabilities of the networks modeled them as subsets of standard first-order logic (FOL), restricted in the operations allowed in order to ensure the tractability that seemed to characterize human reasoning capabilities. The graphical network representations were modeled as providing a visual language for the logic. Sub-sets of FOL targeted on knowledge representation came to be called description logics, and research on these logics has focused on issues of tractability of subsets with differing representation capabilities, and on the implementation of practical inference systems achieving the best possible performance. Semantic network research has kept pace with these developments, providing visual languages for knowledge entry, editing, and presenting the results of inference, that translate unambiguously to the underlying description logics. This paper discusses the design issues for such semantic network formalisms, and illustrates them through detailed examples of significant generic knowledge structures analyzed in the literature, including determinables, contrast sets, genus/differentiae, taxonomies, faceted taxonomies, cluster concepts, family resemblances, graded concepts, frames, definitions, rules, rules with exceptions, essence and state assertions, opposites and contraries, relevance, and so on. Such examples provide important test material for any visual language formalism for logic.

show abstract

Eliciting knowledge and transferring it effectively to a knowledge-based system

Cited by 102 publications

References 21 publications

Challenges to Bayesian decision support using morphological matrices for design: empirical evidence

Challenges to Bayesian decision support using morphological matrices for design: empirical evidence

An iterative approach to build relevant ontology-aware data-driven models

Designing Visual Languages for Description Logics

Contact Info

Product

Resources

About