{"pk":31706,"title":"Analogical Similarity: Performing Structure Alignment in a Connectionist Network","subtitle":null,"abstract":"We describe a connectionist network that performs a com-\nplex, cognitive task. In contrast, the majority of neural net-\nwork research has been devoted to connectionist networks\nthat perform low-level tasks, such as vision. Higher cogni-\ntive tasks, like categorization, analogy, imd similarity may\nultimately rest on alignment of the structured representa-\ntions of two domains. W e model human judgments of simi-\nlarity, as predicted by Structure-Mapping Theory, in the\none-shot mapping task. W e use a localist connectionist\nrepresentation in a Maricov Random Field formalism to\nperform cross-product matching on graph representations\nof propositions. The network performs structured analo-\ngies in its domain flexibly and robustly, resolving local and\nnon-local constraints at multiple levels of abstraction.","language":"eng","license":{"name":"","short_name":"","text":null,"url":""},"keywords":[],"section":"Submitted Presentations","is_remote":true,"remote_url":"https://escholarship.org/uc/item/6ct515f5","frozenauthors":[{"first_name":"Jonathan","middle_name":"B.","last_name":"Handler","name_suffix":"","institution":"The Institute for the Learning Sciences , Evanston","department":""},{"first_name":"Paul","middle_name":"R.","last_name":"Cooper","name_suffix":"","institution":"The Institute for the Learning Sciences , Evanston","department":""}],"date_submitted":null,"date_accepted":null,"date_published":"1993-01-01T18:00:00Z","render_galley":null,"galleys":[{"label":"PDF","type":"pdf","path":"https://journalpub.escholarship.org/cognitivesciencesociety/article/31706/galley/22774/download/"}]}