COMPUTATIONAL ANALYSIS OF CHILD LANGUAGE TRANSCRIPT DATA

NIH RePORTER · NIH · R01 · $179,455 · view on reporter.nih.gov ↗

Abstract

The CHILDES Project provides three crucial tools now being used by language researchers around the world. The first tool is the CHILDES database of computerized transcripts from interactions between children and their parent. This database includes data from earlier funded projects studying children from a variety of social groups, ages, languages, and medical conditions. These data are made available to all researchers. The second major tool is the CHAT transcription format which is used to encode the current database and new data being entered into the system. The third major tool is the CLAN package of data analysis programs which can be used to automatically search the entire database for lexical, syntactic, and discourse patterns relevant to key questions in developmental theory. For the next funding period, the specific aims of the CHILDES project involve five major initiatives that seek to expand and refine the shapes of these three tools. 1. The database initiative. In this area, the goal is to improve the quality and quantity of the existing CHILDES database through additional data entry, detailed documentation, computer-assisted checking, and reformatting. 2. The Johnny Appleseed initiative. The goal of this initiative is to make all three CHILDES tools maximally available to language researchers around the world. 3. The user-friendliness initiative. Here the goal is increase the user-friendliness of the CLAN programs for UNIX, PC-DOS, and Macintosh by bundling the programs into a single menu- driven application for all three systems. 4. The discourse initiative. The goal here is to create a set of CLAN programs that will facilitate study of narrative and conversational discourse patterns. 5. The morphosyntactic initiative. Here the goal is to expand the e- morph and LAG systems to provide automatic morphological analyses and computer-assisted part-of-speech tagging and synaptic parsing of sentences.

Key facts

NIH application ID
2199024
Project number
5R01HD023998-07
Recipient
CARNEGIE-MELLON UNIVERSITY
SAM.gov UEI
U3NKNFLNQ613
Principal Investigator
BRIAN MACWHINNEY
Activity code
R01
Funding institute
NIH
Fiscal year
1994
Award amount
$179,455
Award type
5
Project period
1988-02-01 → 1995-08-31