Datasets:
Tasks:
Text Classification
Modalities:
Text
Formats:
text
Sub-tasks:
intent-classification
Languages:
English
Size:
10K - 100K
License:
Commit History
-
Rename BARBARA 02;04.09.txt to 2- 02;04.09.txt
Verified signature25999c7 -
Rename BARBARA 02;01.92.txt to 1 - 02;01.92.txt
Verified signaturebd09e00 -
Update README.md
Verified signaturece76bdd
-
Update README.md
Verified signatured7f1a72 -
Update README.md
Verified signaturef478a26
-
Upload 85 files
Verified signatureThese are the training files used for MAMAv1, an intent recognition system designed for robotic interaction with children aged 1–4 years. The dataset is derived from the CHILDES database, specifically the Eng-UK collection, including all browsable corpora from Belfast (Barbara, Conor, Courtney, David, John, Michelle, Rachel, Stuart) and additional files from the Fletcher collection (3) (e.g., nwgar, nwjenb, nwkat, nwkery, nwkir, and others). All child utterances were manually extracted and annotated using the INCA-A coding framework (Ninio et al., 1994). Labelling was done using our own categories built on the framework, the framework coding system appeared to extensive with over 60 classes. Annotation was performed by two experienced professionals: a Developmental Psychologist and a Preschool Teacher, ensuring high-quality, developmentally grounded intent labels.f2f39e2 -
initial commit
Verified signaturecbdbe8d