Science of Computation and Language

Find words from age and proportion of acquisition

- Creating child language development database by checklist method -

Abstract

Less is known about language acquisition mechanism because it is difficult to efficiently collect data on vocabulary development from parent-child pairs. To clarify language acquisition mechanism, we collected large scale data on early vocabulary development using vocabulary checklist application. Parents were asked to check whether their child comprehended/produced each word (a total of 2,700 words). Based on the data from 1,300 parent-child pairs, we created child language database including word acquisition properties that were estimated by logistic regression. The database system makes it possible to find the corresponding words from the information about the age and proportion of acquisition. We are now investigating whether the database system is useful for the fields of early education (e.g., picture book search) and pediatric care (e.g., speech therapy).

Photos

Poster


Please click the thumbnail image to open the full-size PDF file.

Presenters


Tessei Kobayashi
Tessei Kobayashi
Innovative Communication Laboratory
Masafumi Matsuda
Masafumi Matsuda
Innovative Communication Laboratory

Yuko Okumura
Yuko Okumura
Innovative Communication Laboratory