Proceedings of the 25th Conference on Computational Natural Language Learning 2021
DOI: 10.18653/v1/2021.conll-1.31
|View full text |Cite
|
Sign up to set email alerts
|

Modeling the Interaction Between Perception-Based and Production-Based Learning in Children’s Early Acquisition of Semantic Knowledge

Abstract: Children learn the meaning of words and sentences in their native language at an impressive speed and from highly ambiguous input. To account for this learning, previous computational modeling has focused mainly on the study of perception-based mechanisms like cross-situational learning. However, children do not learn only by exposure to the input. As soon as they start to talk, they practice their knowledge in social interactions and they receive feedback from their caregivers. In this work, we propose a mode… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
11
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
3
3
2
1

Relationship

0
9

Authors

Journals

citations
Cited by 11 publications
(11 citation statements)
references
References 45 publications
0
11
0
Order By: Relevance
“…One important future research direction is to investigate the implications of flexible role-taking in early interactions (or lack thereof) on the effectiveness of social learning (E. V. Clark, 2022;Kurkul & Corriveau, 2018;Nikolaus & Fourtassi, 2021Yu et al, 2019).…”
Section: Discussionmentioning
confidence: 99%
“…One important future research direction is to investigate the implications of flexible role-taking in early interactions (or lack thereof) on the effectiveness of social learning (E. V. Clark, 2022;Kurkul & Corriveau, 2018;Nikolaus & Fourtassi, 2021Yu et al, 2019).…”
Section: Discussionmentioning
confidence: 99%
“…Such an endeavor will eventually allow us to answer crucial, lingering scientific questions about the universals of human linguistic coordination development and their culture-specific instantiation. The fruit of this effort will be highly impactful to many other domains of cognitive development, the most direct being language learning [7,27,26] but also the acquisition of new knowledge more generally [9,4,14]. Further, adopting Machine Learning tools to model child coordination would help translate our accumulated scientific understanding relatively easily into improvement in child-oriented AI [19,33], thus bridging theory and practice.…”
Section: Discussionmentioning
confidence: 99%
“…One way to do this would be to tie the output layer's weights to input embedding, a well known concept in language modeling (Mikolov et al, 2013;Raffel et al, 2020). Some experiments already implement role alternation, e. g., in multi-agent communication with given language descriptions (Graesser et al, 2019), or in language acquisition from image captions where agents simultaneously learn by cognition and production (Nikolaus & Fourtassi, 2021). We suggest role alternation should also implemented in emergent communication experiments to ensure more linguistically plausible dynamics.…”
Section: Role Alternationmentioning
confidence: 98%