Modeling Probabilistic Reduction using Information Theory and Naive Discriminative Learning

TL;DR

This paper compares information-theoretic probabilistic models with Naive Discriminative Learning for modeling acoustic word duration, finding N-gram models outperform NDL but that integrating information formulas into NDL enhances its performance.

Contribution

It introduces a novel integration of information-theoretic formulas into NDL, improving its effectiveness in modeling acoustic reduction compared to traditional NDL.

Findings

N-gram models outperform NDL models in predicting acoustic reduction.

Incorporating information-theoretic formulas into NDL improves its predictive performance.

Combining frequency, contextual predictability, and average predictability is crucial for modeling acoustic reduction.

Abstract

This study compares probabilistic predictors based on information theory with Naive Discriminative Learning (NDL) predictors in modeling acoustic word duration, focusing on probabilistic reduction. We examine three models using the Buckeye corpus: one with NDL-derived predictors using information-theoretic formulas, one with traditional NDL predictors, and one with N-gram probabilistic predictors. Results show that the N-gram model outperforms both NDL models, challenging the assumption that NDL is more effective due to its cognitive motivation. However, incorporating information-theoretic formulas into NDL improves model performance over the traditional model. This research highlights a) the need to incorporate not only frequency and contextual predictability but also average contextual predictability, and b) the importance of combining information-theoretic metrics of predictability and information derived from discriminative learning in modeling acoustic reduction.