×

How to build a knowledge space by querying an expert. (English) Zbl 0725.92029

The authors are concerned with the problem of building knowledge structures - KS - for particular domains (a KS is defined as the collection of all possible knowledge states, while the knowledge state of an individual is formalized as the subset of notions he has mastered; finally, a notion is identified with an equivalence class of questions/problems, testing just that notion). The algebraic foundation of this approach was investigated by J.-P. Doignon and J.-C. Falmagne [Int. J. Man-Mach. Stud. 23, 175-196 (1985; Zbl 0581.68066)]. If the KS is closed under union and intersection, then it can be equivalently specified by a quasi-order on the set of problems. When the - less realistic - assumption of closure under intersection is dropped, a representation of KSs by surmise systems - a variant of AND/OR graphs - is possible. The KSs which are closed under union are called knowledge spaces (KSP).
The point of the current paper is to derive an alternative representation for KSP. Here, quasi-orders are used as relations on the power set of the set of problems; this representation is at the basis of a procedure that translates the responses of an expert to a set of queries of a specific form into a corresponding KSP. Birkhoff’s theorem [G. Birkhoff, Duke Math. J. 3, 443-454 (1937; Zbl 0017.19403)], which plays a central role in the quasi-order representation, can be derived from a Galois connection between the collection of KSs and the collection of binary relations on the set of problems. The main result of the paper under review is an extension of Birkhoff’s theorem by establishing a more general Galois connection between KSs and relations on the power set of the set of problems.

MSC:

91E10 Cognitive psychology
06A15 Galois correspondences, closure operators (in relation to ordered sets)
68T30 Knowledge representation
06A06 Partial orders, general
68T05 Learning and adaptive systems in artificial intelligence
91E40 Memory and learning in psychology
PDF BibTeX XML Cite
Full Text: DOI

References:

[1] Birkoff, G., Rings of sets, Duke mathematical journal, 3, 443-454, (1937)
[2] Birkhoff, G., Lattice theory, () · Zbl 0126.03801
[3] Burrigana, L., Organization by rules in finite sequences, (1988), Submitted for publication
[4] Degreef, E.; Doignon, J.-P.; Ducamp, A.; Falgmagne, J.-C., Languages for the assessment of knowledge, Journal of mathematical psychology, 30, 243-256, (1986) · Zbl 0609.92037
[5] Doignon, J.-P.; Falmagne, J.-C., Spaces for the assessment f knowledge, International journal of man-machine studies, 23, 175-196, (1985) · Zbl 0581.68066
[6] Falmagne, J.-C., A latent trait theory via a stochastic learning theoyr for a knowledge space, Psychometrika, 54, 283-303, (1990)
[7] Falmagne, J.-C.; Doignon, J.-P., A class of stochastic procedures for the assessment of knowledge, British journal of statistical and mathematical psychology, 41, 1-23, (1988) · Zbl 0719.92026
[8] Falmagne, J.-C.; Doignon, J.-P., A Markovian procedure for assessing the state of a system, Journal of mathematical psychology, 32, 232-258, (1988) · Zbl 0654.92020
[9] Kambouri, M.; Koppen, M.; Villano, M.; Falmagne, J.-C., Knowledge assessment: tapping human expertise, (1989), In progress
[10] Koppen, M., Extracting human expertise for constructing knowledge spaces: an algorithm, (1989), Submitted for publication · Zbl 0768.92029
[11] Monjardet, B., Tresses, fuseaux, préordres et topologies, Mathématiques et sciences humaines, 30, 11-22, (1970)
[12] Muller, C., A procedure to facilitate an Expert’s judgements on a set of rules, (), in press
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.