Conditional and joint models for grapheme-to-phoneme conversion
Conditional and joint models for grapheme-to-phoneme conversion
复制标题
DOI：
10.21437/eurospeech.2003-584
复制
发表时间：
2003-09
期刊：
影响因子：
--
通讯作者：
Stanley F. Chen
中科院分区：
文献类型：
--
作者：
Stanley F. Chen
在这项工作中，我们介绍了几种字素到音素转换的模型：条件最大熵模型，联合最大熵n元语法模型，以及结合音节fi阳离子的联合最大熵n元语法模型。我们考察了条件模型和联合模型对这项任务的相对优点，fi发现联合模型有许多优势。我们表明，我们的最好模型-联合n-gram模型的性能与文献中报道的英语字素到音素转换的最佳结果相比是有利的，有时差距很大。在本文的后半部分，我们考虑了融合不同音素集合的发音词汇的任务。我们表明，字素到音素转换的模型可以有效地适应这一任务。
In this work, we introduce several models for grapheme-to-phoneme conversion: a conditional maximum entropy model, a joint maximum entropy n -gram model, and a joint maximum entropy n -gram model with syllabification. We examine the relative merits of conditional and joint models for this task, and find that joint models have many advantages. We show that the performance of our best model, the joint n -gram model, compares favorably with the best results for English grapheme-to-phoneme conversion reported in the literature, sometimes by a wide margin. In the latter part of this paper, we consider the task of merging pronunciation lexicons expressed in different phone sets. We show that models for grapheme-to-phoneme conversion can be adapted effectively to this task.