Abstract

Speech contains all the lexical content of text in addition to other features like pitch and prosody. Could this type of non-lexical information be useful for tasks like summarization? I will discuss recent work in speech summarization and present a plan to deduce the utility of speech in automatic summarization. Feedback on this work in progress is highly appreciated.

Bio

Kaavya Chaparala is a first year Computer Science PhD student advised by Najim Dehak.