A Comparative Study of Mixture Models for Automatic Topic Segmentation of Multiparty Dialogues
Maria Georgescul, Alexander Clark, Susan Armstrong · International Joint Conference on Natural Language Processing · 2008
In this article we address the task of automatic text structuring into linear and nonoverlapping thematic episodes at a coarse level of granularity. In particular, we deal with topic segmentation on multi-party meeting recording transcripts, which pose specific challenges for topic segmentation models. We present a comparative study of two probabilistic mixture models. Based on lexical features, we use these models in parallel in order to generate a low dimensional input representation for topic segmentation. Our experiments demonstrate that in this manner important information is captured from the data through less features.