Confusion2vec 2.0: Enriching Ambiguous Spoken Language Representations with Subwords

TL;DR

This paper introduces Confusion2vec 2.0, a novel word embedding that incorporates subword information to better capture acoustic ambiguities in spoken language, improving performance on various linguistic tasks with ASR errors.

Contribution

It proposes an unsupervised subword-based extension of Confusion2vec that effectively models acoustic ambiguities directly from ASR lattices, outperforming existing embeddings.

Findings

Significantly better performance on semantic, syntactic, and acoustic analogy tasks.

Enhanced spoken language intent detection accuracy.

Reduces need for retraining NLU models on ASR transcripts.

Abstract

Word vector representations enable machines to encode human language for spoken language understanding and processing. Confusion2vec, motivated from human speech production and perception, is a word vector representation which encodes ambiguities present in human spoken language in addition to semantics and syntactic information. Confusion2vec provides a robust spoken language representation by considering inherent human language ambiguities. In this paper, we propose a novel word vector space estimation by unsupervised learning on lattices output by an automatic speech recognition (ASR) system. We encode each word in confusion2vec vector space by its constituent subword character n-grams. We show the subword encoding helps better represent the acoustic perceptual ambiguities in human spoken language via information modeled on lattice structured ASR output. The usefulness of the proposed Confusion2vec representation is evaluated using semantic, syntactic and acoustic analogy and word similarity tasks. We also show the benefits of subword modeling for acoustic ambiguity representation on the task of spoken language intent detection. The results significantly outperform existing word vector representations when evaluated on erroneous ASR outputs. We demonstrate that Confusion2vec subword modeling eliminates the need for retraining/adapting the natural language understanding models on ASR transcripts.