{"pk":25706,"title":"Cross-situational cues are relevant for early word segmentation","subtitle":null,"abstract":"Existing models of infant word learning have mainly\nassumed that the learner is capable of segmenting words from\nspeech before grounding them to their referential meaning,\nwhile segmentation itself has been treated relatively\nindependently of meaning acquisition. In this paper, we argue\nthat situated cues such as visually perceived concrete objects\nor actions are not just important for word-to-meaning\nmapping, but that they are useful in pre-linguistic word\nsegmentation, thereby helping the learner to bootstrap the\nlanguage learning process. We present a model where joint\nacquisition of proto-lexical segments and their meanings\nmaximizes the referential quality of the lexicon, and where\nlearning can occur without any a priori knowledge of the\nlanguage or its linguistically relevant units. We investigate the\nbehavior of the model using a computational implementation\nof statistical learning, showing successful word segmentation\nunder varying degrees of referential uncertainty.","language":"eng","license":{"name":"","short_name":"","text":null,"url":""},"keywords":[{"word":"word learning; segmentation; meaning\nacquisition; computational modeling; synergies in word\nlearning; language acquisition"}],"section":"Papers","is_remote":true,"remote_url":"https://escholarship.org/uc/item/336868mc","frozenauthors":[{"first_name":"Okko","middle_name":"","last_name":"Rasanen","name_suffix":"","institution":"Aalto University","department":""},{"first_name":"Heikki","middle_name":"","last_name":"Rasilo","name_suffix":"","institution":"Aalto University","department":""}],"date_submitted":null,"date_accepted":null,"date_published":"2015-01-01T18:00:00Z","render_galley":null,"galleys":[{"label":"PDF","type":"pdf","path":"https://journalpub.escholarship.org/cognitivesciencesociety/article/25706/galley/15330/download/"}]}