<span xss=removed>为了提高语音查询项检索系统中集外词检索的性能</span><span xss=removed>,</span><span xss=removed>在加权有限状态转换器（</span><span xss=removed>weighted finite-state transducer, WFST</span><span xss=removed>）框架下提出了一种基于音素混淆模型的集外词查询项扩展技术</span><span xss=removed>,</span><span style="font-family:宋体;fo
