Morphological Segmentation for Keyword Spotting
Morphological Segmentation for Keyword Spotting
复制标题
DOI：
10.3115/v1/d14-1095
复制
发表时间：
2014-10
期刊：
影响因子：
--
通讯作者：
Karthik Narasimhan;Damianos G. Karakos;R. Schwartz;Stavros Tsakalidis;R. Barzilay
中科院分区：
文献类型：
--
作者：
Karthik Narasimhan;Damianos G. Karakos;R. Schwartz;Stavros Tsakalidis;R. Barzilay
我们探讨了形态分割对关键词定位(KWS)的影响。尽管有潜在的好处，最先进的KWS系统并不使用形态信息。在本文中，我们使用监督和非监督形态切分得到的子词单元来扩充最新的KWS系统，并与语音切分和音节切分进行了比较。实验表明，语素提高了KWS系统的整体性能。然而，在KWS中使用音节单位时，音节单位的表现与形态单位不相上下。通过将形态切分、语音切分和音节切分结合起来，我们展示了显著的性能提升。
We explore the impact of morphological segmentation on keyword spotting (KWS). Despite potential benefits, stateof-the-art KWS systems do not use morphological information. In this paper, we augment a state-of-the-art KWS system with sub-word units derived from supervised and unsupervised morphological segmentations, and compare with phonetic and syllabic segmentations. Our experiments demonstrate that morphemes improve overall performance of KWS systems. Syllabic units, however, rival the performance of morphological units when used in KWS. By combining morphological, phonetic and syllabic segmentations, we demonstrate substantial performance gains.