In this study we compare two sequence learning approaches to chunking 
turninternal dialogue act sequences. These assign a dialogue act label to 
each token in the transcribed speech stream of a dialogue participant, 
additionally guessing if the token is at the beginning of, inside, or 
outside that specific dialogue act. 

Experimental findings show that both our approaches – conditional random fields
and memory-based tagging – largely improve over local classification methods. 
We discuss the interplay between turn-taking transcription granularity and
dialogue act chunking.