Mining Word Boundaries from Speech-Text Parallel Data for Cross-domain Chinese Word Segmentation

TL;DR

This paper introduces a novel method to mine word boundaries from speech-text parallel data using forced alignment and pause analysis, enhancing Chinese Word Segmentation across domains.

Contribution

It is the first to explicitly extract word boundaries from speech-text data for Chinese Word Segmentation, employing a probability-based filtering and a complete-then-train strategy.

Findings

Effective cross-domain CWS performance improvement

Successful extraction of word boundaries from speech data

Robust training strategy enhances segmentation accuracy

Abstract

Inspired by early research on exploring naturally annotated data for Chinese Word Segmentation (CWS), and also by recent research on integration of speech and text processing, this work for the first time proposes to explicitly mine word boundaries from speech-text parallel data. We employ the Montreal Forced Aligner (MFA) toolkit to perform character-level alignment on speech-text data, giving pauses as candidate word boundaries. Based on detailed analysis of collected pauses, we propose an effective probability-based strategy for filtering unreliable word boundaries. To more effectively utilize word boundaries as extra training data, we also propose a robust complete-then-train (CTT) strategy. We conduct cross-domain CWS experiments on two target domains, i.e., ZX and AISHELL2. We have annotated about 1,000 sentences as the evaluation data of AISHELL2. Experiments demonstrate the effectiveness of our proposed approach.