| Title | Learn It First: Grounding Language in Compositional Event-Predictive Encodings |
| Publication Type | Conference Paper |
| Year of Publication | 2021 |
| Authors | Stegemann-Philipps, Christian, Butz, Martin V. |
| Conference Name | 2021 IEEE International Conference on Development and Learning (ICDL) |
| Date Published | aug |
| Keywords | cognitive architecture, compositionality, Conferences, encoding, Event-Cognition, Grounding, Image resolution, Language Grounding, learning (artificial intelligence), mental simulation, Natural languages, Pediatrics, pubcrawl |
| Abstract | While language learning in infants and toddlers progresses somewhat seamlessly, in artificial systems the grounding of language in knowledge structures that are learned from sensorimotor experiences remains a hard challenge. Here we introduce LEARNA, which learns event-characterizing abstractions to resolve natural language ambiguity. LEARNA develops knowledge structures from simulated sensorimotor experiences. Given a possibly ambiguous descriptive utterance, the learned knowledge structures enable LEARNA to infer environmental scenes, and events unfolding within, which essentially constitute plausible imaginations of the utterance's content. Similar event-predictive structures may help in developing artificial systems that can generate and comprehend descriptions of scenes and events. |
| DOI | 10.1109/ICDL49984.2021.9515621 |
| Citation Key | stegemann-philipps_learn_2021 |
Learn It First: Grounding Language in Compositional Event-Predictive Encodings