Minimum cut model for spoken lecture segmentation
Author(s)
Malioutov, Igor (Igor Mikhailovich)
DownloadFull printable version (17.81Mb)
Other Contributors
Massachusetts Institute of Technology. Dept. of Electrical Engineering and Computer Science.
Advisor
Regina Barzilay.
Terms of use
Metadata
Show full item recordAbstract
We introduce a novel unsupervised algorithm for text segmentation. We re-conceptualize text segmentation as a graph-partitioning task aiming to optimize the normalized-cut criterion. Central to this framework is a contrastive analysis of lexical distribution that simultaneously optimizes the total similarity within each segment and dissimilarity across segments. Our experimental results show that the normalized-cut algorithm obtains performance improvements over the state-of-the-art techniques on the task of spoken lecture segmentation. Another attractive property of the algorithm is robustness to noise. The accuracy of our algorithm does not deteriorate significantly when applied to automatically recognized speech. The impact of the novel segmentation framework extends beyond the text segmentation domain. We demonstrate the power of the model by applying it to the segmentation of raw acoustic signal without intermediate speech recognition.
Description
Thesis (S.M.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, February 2007. Includes bibliographical references (leaves 129-132).
Date issued
2007Department
Massachusetts Institute of Technology. Department of Electrical Engineering and Computer SciencePublisher
Massachusetts Institute of Technology
Keywords
Electrical Engineering and Computer Science.