Text-to-Scene for Field Linguistics

Project Summary

This research aims at creating a novel tool for fieldwork. We call this tool the WordsEye Linguistics Tool, or WELT. WELT is based on WordsEye, an existing text-to-scene tool which has been developed by the PIs and their students. WordsEye allows for the automatic generation of 3D scenes from written input. WELT tool will have two modes of operation:

English input will automatically generate a picture which can be used to elicit a targeted description.
Endangered language input will automatically generate a picture representing the meaning of the input according to a formal grammar of spatial expressions; the picture can then be used to verify the grammar with native speakers.

While we intend to create a tool that will be generally useful, we will initially develop WELT based on scenarios involving Arrernte. Furthermore, we will develop the tool in order to test formal grammars of spatial expressions of two types: (a) associated motion and (b) spatial location.

Sponsorship

This research is funded 2012-2013 by the National Science Foundation, Division of Information & Intelligent Systems, award #1160700 "Using Computational Tools to Facilitate Corpus Collection and Language Use in Arrernte."