Science and Technology Production
Congress
Authorship
RICARDO MINERVINO
;
NICOLAS OBERHOLZER
;
TRENCH, JUAN MAXIMO
Date
2008
Publishing House and Editing Place
Cognitive Science Society
ISSN
978-0-9768318-4-6
Summary
Information provided by the agent in
SIGEVA
Most computational models of analogical mapping and evaluation of analogical relatedness (e.g., SME, ACME and LISA) reduce the effects of semantics on these processes to the influence of similarities between propositional elements to be paired. Two experiments were carried out to show that people do not always follow this kind of similarities. In Experiment 1, when comparing a single proposition base analog with two alternative target analogs, participants judged as more analogous those that di...
Most computational models of analogical mapping and evaluation of analogical relatedness (e.g., SME, ACME and LISA) reduce the effects of semantics on these processes to the influence of similarities between propositional elements to be paired. Two experiments were carried out to show that people do not always follow this kind of similarities. In Experiment 1, when comparing a single proposition base analog with two alternative target analogs, participants judged as more analogous those that did not share that type of similarity. In Experiment 2 participants solved ambiguous mappings between propositions framed within systems of relations and tasks of cause identification. They favored matchings between propositions lacking element to element similarity. The implications of these results for computational models of analogical thinking are discussed
Show more
Show less
Key Words
SIMILARITYANALOGYREREPRESENTATIONMAPPING