Like {\tt BATCH\_DIALOGUE\_SPEECH\_AGAIN}, but uses the version of the speech corpus tagged {\tt Arg1}. Input is taken from the transcriptions file specified by the config parameter \begin{verbatim} dialogue_speech_corpus(Arg1) \end{verbatim} and output is written to the file specified by the config parameter \begin{verbatim} dialogue_speech_corpus_results(Arg1) \end{verbatim}

The config file also needs to define all the entries associated with spoken dialogue applications.