非语言特征建模的普通话新闻语音摘要研究

首次研究对使用非语言特征进行普通话广播新闻摘要的建模方法。评估语音特征、语言特征、结构特征等对抽取摘要的贡献。结果表明,仅用语音特征和结构特征这两大类与语言特征无关的特征,所建立的摘要模型,其摘要抽取性能良好,F-measure达到了0.565。此外,还发现,结构特征要优于语言特征;单独使用声学特征所训练出来的摘要模型,性能也达到了平均F-measure0.391。这些发现使得语音摘要的抽取性能在一定程度上不受语音识别准确率的限制。

Speech Summarization Without Lexical Features for Mandarin Broadcast News

We present the first known empirical study on speech summarization without lexical features for Mandarin broadcast news. We evaluate acoustic, lexical and structural features as predictors of summary sentences. We find that the summarizer yields good performance at the average F-measure of 0.565 even by using the combination of acoustic and structural features alone, which are independent of lexical features. In addition, we show that structural features are superior to lexical features and our summarizer performs surprisingly well at the average F-measure of 0.391 by using only acoustic features. These findings enable us to summarize speech without placing a stringent demand on speech recognition accuracy.