Introduction

About this project

Ontolex: An Extensible, Internet-based Framework for Modeling Second Language Acquisition

Emily Fortuna* and Larry Reeker
National Institute of Standards and Technology
*Also Rice University

Since the birth of Computer Science and modern Linguistics, numerous cognitive computer models have emerged with a myriad of approaches as a method to study the mysterious black box of the human mind. This project is the second generation of an effort to create an extensible reference model for these models, or a meta-model of second language learning. The internet-based framework, called Ontolex, is an effort to abstract ontology design and the ideas of several grammar theories in order to universalize its application and extensibility and eliminate dependency on any specific natural language. Ontolex creates word-to-ontology relationships via a three tier system mapping words to lexemes to concepts in the ontology. This learning system, part of an ongoing implementation, will model three general phases in the human second language learning process: vocabulary building, formulaic speech and translation, and unique utterance generation with user reinforcement feedback.

Microsoft Word: Final Report

This framework is written in user-friendly Python (and XHTML/CSS) , chosen for its conciseness, ease of maintenance, and ease of extensibility. Interested in building off of the framework for your research? Download the source code.

Powerpoint Presentation about Ontolex given during the Summer Undergraduate Research Fellowship presentations at the National Institute of Standards and Technology.