Predicting word learning to boost child language acquisition.

Roxburgh, A., Grasso, F., & Payne, T. (2017). Predicting word learning to boost child language acquisition. Poster presented at the 7th International Conference on Digital Health, DH '17, London, United Kingdom — July 02 - 05, 2017

Abstract

Children’s level of language acquisition from around the age of two years upwards has been shown to be positively correlated with their later performance at school [2]. It follows that one way to improve a child’s future school performance would be to encourage him or her to acquire language as early as possible. Children prefer to learn words that they can categorise with other words that they already know [3,4] - firstly through a similarity of shape and then though other more complex associations as the child’s mind creates more categories. It follows, then, that if a system used by the parent – such as a mobile device based application – could be used to log information about the child's language development, it could also give advice to the parent about which words to encourage the child to learn next, those words having been judged by the system to be the most associated with words in the child’s existing vocabulary. By suggesting the words most likely to be learned by the child in the immediate future, perhaps the efficiency of the child's language acquisition process could be boosted, leading to the child learning language at an earlier age. This could be of particular help to children who are already in groups likely to experience a delay in language acquisition. However, no two children are the same. A child living on a farm may have different environmental influences on their vocabulary compared to a child living in an inner-city area. It follows that two words that may be closely semantically linked in one child’s mind may not be linked at all in the mind of another.
Deep learning techniques combined with cheaper and higher-performance computer hardware have shown great success in moving the field of pattern recognition forward in recent years, and are being used to improve many applications of artificial intelligence.
Our primary aim is to create, using machine learning methods, a computational model that will predict the words that a typically-developing child is most likely to learn next. The work improves on the model in [1] to include also static and time-varying environmental and vocabulary data, and including a prediction time of less than a month in order to be able to provide timely advice to parents. Ultimately, we plan to train the model on atypically developing children such as those with ASD.
We will present our preliminary results in this direction.

[1] N. Beckage, M. Mozer, and E. Colunga. "Predicting a Child's Trajectory of Lexical Acquisition." CogSci. 2015.
[2] D. Bleses, G. Makransky, P. S. Dale, A. Hojen, and B. A. Ari, “Early productive vocabulary predicts academic achievement 10 years later,” Applied Psycholinguistics, pp. 1–16, 2016.
[3]A. Borovsky and J. L. Elman, “Language input and semantic categories: A relation between cognition and early word learning,” Journal of Child Language, vol. 33, no. 4, pp. 759–790, 2006.
[4]G. Diesendruck, “Mechanisms of Word Learning,” in Blackwell Handbook of Language Development, ch. 13, pp. 257–276, 2008.