Spatial cognition of geometric figures in the context of proportional analogies

Autor(en): Schwering, A.
Kühnberger, K.-U. 
Krumnack, U. 
Gust, H.
Stichwörter: Adaptation; Computational model for spatial cognition; Context; Geometric proportional analogy; Re-representation; Re-representation, Information theory, Image registration
Erscheinungsdatum: 2009
Journal: Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen: 5756 LNCS
Startseite: 18
Seitenende: 35
Zusammenfassung: 
The cognition of spatial objects differs among people and is highly influenced by the context in which a spatial object is perceived. We investigated experimentally how humans perceive geometric figures in geometric proportional analogies and discovered that subjects perceive structures within the figures which are suitable for solving the analogy. Humans do not perceive the elements within a figure individually or separately, but cognize the figure as a structured whole. Furthermore, the perception of each figure in the series of analogous figures is influenced by the context of the whole analogy. A computational model which shall reflect human cognition of geometric figures must be flexible enough to adapt the representation of a geometric figure and produce a similarly structured representation as humans do while solving the analogy. Furthermore, it must be able to take into account the context, i.e. structures and transformations in other geometric figures in the analogy. © 2009 Springer Berlin Heidelberg.
Beschreibung: 
Conference of 9th International Conference on Spatial Information Theory, COSIT 2009 ; Conference Date: 21 September 2009 Through 25 September 2009; Conference Code:77840
ISBN: 9783642038310
ISSN: 03029743
DOI: 10.1007/978-3-642-03832-7_2
Externe URL: https://www.scopus.com/inward/record.uri?eid=2-s2.0-70350412309&doi=10.1007%2f978-3-642-03832-7_2&partnerID=40&md5=e7fb5d94b3da2208be8b917fc48b9c4a

Zur Langanzeige

Google ScholarTM

Prüfen

Altmetric