A perceptual criterion for visually controlling learning

Masaki Suwa, Hiroshi Motoda

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Acquiring search control knowledge of high utility is essential to reasoners in speeding up their problem-solving performance. In the domain of geometry problem-solving, the role of “perceptual chunks”, an assembly of diagram elements many problems share in common, in effectively guiding problem-solving search has been extensively studied, but the issue of learning these chunks from experiences has not been addressed so far. Although the explanation-based learning technique is a typical learner for search control knowledge, the goal-orientedness of its chunking criterion leads to produce such search control knowledge that can only be used for directly accomplishing a target-concept, which is totally different from what perceptual-chunks are for. This paper addresses the issues of acquiring domain-specific perceptual-chunks and demonstrating the utility of acquired chunks. The proposed technique is that the learner acquires, for each control decision node in the problem solving traces, a chunk which is an assembly of diagram elements that can be visually recognizable and grouped together with the control decision node. Recognition rules implement this chunking criterion in the learning system PCLEARN. We show the feasibility of the proposed technique by investigating the applicability and cost-effective utility of the learned perceptual chunks in the geometry domain.

Original languageEnglish
Title of host publicationAlgorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings
PublisherSpringer Verlag
Pages356-369
Number of pages14
Volume744 LNAI
ISBN (Print)9783540573708
Publication statusPublished - 1993
Externally publishedYes
Event4th Workshop on Algorithmic Learning Theory, ALT 1993 - Tokyo, Japan
Duration: 1993 Nov 81993 Nov 10

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume744 LNAI
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Other

Other4th Workshop on Algorithmic Learning Theory, ALT 1993
CountryJapan
CityTokyo
Period93/11/893/11/10

Fingerprint

Diagram
Geometry
Learning Systems
Vertex of a graph
Learning systems
Trace
Learning
Target
Costs
Knowledge
Concepts
Experience

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Cite this

Suwa, M., & Motoda, H. (1993). A perceptual criterion for visually controlling learning. In Algorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings (Vol. 744 LNAI, pp. 356-369). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 744 LNAI). Springer Verlag.

A perceptual criterion for visually controlling learning. / Suwa, Masaki; Motoda, Hiroshi.

Algorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings. Vol. 744 LNAI Springer Verlag, 1993. p. 356-369 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 744 LNAI).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Suwa, M & Motoda, H 1993, A perceptual criterion for visually controlling learning. in Algorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings. vol. 744 LNAI, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 744 LNAI, Springer Verlag, pp. 356-369, 4th Workshop on Algorithmic Learning Theory, ALT 1993, Tokyo, Japan, 93/11/8.
Suwa M, Motoda H. A perceptual criterion for visually controlling learning. In Algorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings. Vol. 744 LNAI. Springer Verlag. 1993. p. 356-369. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Suwa, Masaki ; Motoda, Hiroshi. / A perceptual criterion for visually controlling learning. Algorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings. Vol. 744 LNAI Springer Verlag, 1993. pp. 356-369 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{cedcd24907e14824920eda9942a5dcb0,
title = "A perceptual criterion for visually controlling learning",
abstract = "Acquiring search control knowledge of high utility is essential to reasoners in speeding up their problem-solving performance. In the domain of geometry problem-solving, the role of “perceptual chunks”, an assembly of diagram elements many problems share in common, in effectively guiding problem-solving search has been extensively studied, but the issue of learning these chunks from experiences has not been addressed so far. Although the explanation-based learning technique is a typical learner for search control knowledge, the goal-orientedness of its chunking criterion leads to produce such search control knowledge that can only be used for directly accomplishing a target-concept, which is totally different from what perceptual-chunks are for. This paper addresses the issues of acquiring domain-specific perceptual-chunks and demonstrating the utility of acquired chunks. The proposed technique is that the learner acquires, for each control decision node in the problem solving traces, a chunk which is an assembly of diagram elements that can be visually recognizable and grouped together with the control decision node. Recognition rules implement this chunking criterion in the learning system PCLEARN. We show the feasibility of the proposed technique by investigating the applicability and cost-effective utility of the learned perceptual chunks in the geometry domain.",
author = "Masaki Suwa and Hiroshi Motoda",
year = "1993",
language = "English",
isbn = "9783540573708",
volume = "744 LNAI",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "356--369",
booktitle = "Algorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings",
address = "Germany",

}

TY - GEN

T1 - A perceptual criterion for visually controlling learning

AU - Suwa, Masaki

AU - Motoda, Hiroshi

PY - 1993

Y1 - 1993

N2 - Acquiring search control knowledge of high utility is essential to reasoners in speeding up their problem-solving performance. In the domain of geometry problem-solving, the role of “perceptual chunks”, an assembly of diagram elements many problems share in common, in effectively guiding problem-solving search has been extensively studied, but the issue of learning these chunks from experiences has not been addressed so far. Although the explanation-based learning technique is a typical learner for search control knowledge, the goal-orientedness of its chunking criterion leads to produce such search control knowledge that can only be used for directly accomplishing a target-concept, which is totally different from what perceptual-chunks are for. This paper addresses the issues of acquiring domain-specific perceptual-chunks and demonstrating the utility of acquired chunks. The proposed technique is that the learner acquires, for each control decision node in the problem solving traces, a chunk which is an assembly of diagram elements that can be visually recognizable and grouped together with the control decision node. Recognition rules implement this chunking criterion in the learning system PCLEARN. We show the feasibility of the proposed technique by investigating the applicability and cost-effective utility of the learned perceptual chunks in the geometry domain.

AB - Acquiring search control knowledge of high utility is essential to reasoners in speeding up their problem-solving performance. In the domain of geometry problem-solving, the role of “perceptual chunks”, an assembly of diagram elements many problems share in common, in effectively guiding problem-solving search has been extensively studied, but the issue of learning these chunks from experiences has not been addressed so far. Although the explanation-based learning technique is a typical learner for search control knowledge, the goal-orientedness of its chunking criterion leads to produce such search control knowledge that can only be used for directly accomplishing a target-concept, which is totally different from what perceptual-chunks are for. This paper addresses the issues of acquiring domain-specific perceptual-chunks and demonstrating the utility of acquired chunks. The proposed technique is that the learner acquires, for each control decision node in the problem solving traces, a chunk which is an assembly of diagram elements that can be visually recognizable and grouped together with the control decision node. Recognition rules implement this chunking criterion in the learning system PCLEARN. We show the feasibility of the proposed technique by investigating the applicability and cost-effective utility of the learned perceptual chunks in the geometry domain.

UR - http://www.scopus.com/inward/record.url?scp=85029438649&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85029438649&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85029438649

SN - 9783540573708

VL - 744 LNAI

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 356

EP - 369

BT - Algorithmic Learning Theory - 4th International Workshop, ALT 1993, Proceedings

PB - Springer Verlag

ER -