arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1807.00543cs.CL
收藏

会话语音的标点预测模型

Punctuation Prediction Model for Conversational Speech

Intelligent Wire(智能有线公司);AGH University of Science and Technology(AGH科技大学);Wrocław University of Technology(弗罗茨瓦夫理工大学);The Johns Hopkins University(约翰斯·霍普金斯大学)

Piotr Żelasko, Piotr Szymański, Jan Mizgajski, Adrian Szymczak, Yishay Carmiel, Najim Dehak

摘要:

ASR系统通常不预测任何标点符号或大写字母。缺少标点符号会导致结果呈现出现问题,并使人类读者和现成的自然语言处理算法感到困惑。为了克服这些限制,我们训练了深度神经网络(DNN)序列标记模型的两个变体-双向长短期记忆(BLSTM)和卷积神经网络(CNN),以预测标点符号。模型在Fisher语料库上训练,其中包括标点符号注释。在我们的实验中,我们结合使用序列比对算法的时间对齐和标点Fisher语料库成绩单。神经网络在与会话侧指示符和单词时间信息对齐的Fisher转录中的单词的通用网络爬行GloVe嵌入上进行训练。CNN产生更好的精度,BLSTM往往具有更好的召回率。虽然BLSTM总体上犯的错误更少,但CNN预测的标点符号更准确--尤其是在问号的情况下。我们的研究结果构成了重要的证据,表明单词在时间上的分布以及预先训练的嵌入在标点符号预测任务中是有用的。

英文摘要:

An ASR system usually does not predict any punctuation or capitalization. Lack of punctuation causes problems in result presentation and confuses both the human reader andoff-the-shelf natural language processing algorithms. To overcome these limitations, we train two variants of Deep Neural Network (DNN) sequence labelling models - a Bidirectional Long Short-Term Memory (BLSTM) and a Convolutional Neural Network (CNN), to predict the punctuation. The models are trained on the Fisher corpus which includes punctuation annotation. In our experiments, we combine time-aligned and punctuated Fisher corpus transcripts using a sequence alignment algorithm. The neural networks are trained on Common Web Crawl GloVe embedding of the words in Fisher transcripts aligned with conversation side indicators and word time infomation. The CNNs yield a better precision and BLSTMs tend to have better recall. While BLSTMs make fewer mistakes overall, the punctuation predicted by the CNN is more accurate - especially in the case of question marks. Our results constitute significant evidence that the distribution of words in time, as well as pre-trained embeddings, can be useful in the punctuation prediction task.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑