5th International Conference on Spoken Language Processing
The segmentation of text and speech into topics and subtopics is an important step in document interpretation. For text, formatting information, such as headings and paragraphing, is available to aid in this endeavor, although this information is by no means sufficient. For speech, the task is even more difficult. We present results of the application of machine learning techniques to the automatic identification of intonational phrases beginning and ending 'topics' determined independently by annotators for two corpora | the Boston Directions Corpus and the Broadcast News (HUB-4) DARPA/NIST database.
Bibliographic reference. Hirschberg, Julia / Nakatani, Christine H. (1998): "Acoustic indicators of topic segmentation", In ICSLP-1998, paper 0976.