Bridging the Gap: An Intermediate Language for Enhanced and Cost-Effective Grapheme-to-Phoneme Conversion with Homographs with Multiple Pronunciations Disambiguation

TL;DR

This paper presents a novel intermediate language and a hybrid approach combining LLM prompting and sequence-to-sequence models to improve Persian G2P conversion, especially for homographs with multiple pronunciations, achieving state-of-the-art accuracy.

Contribution

It introduces an intermediate language tailored for Persian, integrating LLM prompts and a specialized transliteration architecture to disambiguate homographs and enhance G2P accuracy.

Findings

Significant reduction in Phoneme Error Rate compared to previous methods

Effective disambiguation of homographs with multiple pronunciations

Benchmark performance surpassing existing state-of-the-art approaches

Abstract

Grapheme-to-phoneme (G2P) conversion for Persian presents unique challenges due to its complex phonological features, particularly homographs and Ezafe, which exist in formal and informal language contexts. This paper introduces an intermediate language specifically designed for Persian language processing that addresses these challenges through a multi-faceted approach. Our methodology combines two key components: Large Language Model (LLM) prompting techniques and a specialized sequence-to-sequence machine transliteration architecture. We developed and implemented a systematic approach for constructing a comprehensive lexical database for homographs with multiple pronunciations disambiguation often termed polyphones, utilizing formal concept analysis for semantic differentiation. We train our model using two distinct datasets: the LLM-generated dataset for formal and informal Persian and the B-Plus podcasts for informal language variants. The experimental results demonstrate superior performance compared to existing state-of-the-art approaches, particularly in handling the complexities of Persian phoneme conversion. Our model significantly improves Phoneme Error Rate (PER) metrics, establishing a new benchmark for Persian G2P conversion accuracy. This work contributes to the growing research in low-resource language processing and provides a robust solution for Persian text-to-speech systems and demonstrating its applicability beyond Persian. Specifically, the approach can extend to languages with rich homographic phenomena such as Chinese and Arabic