This command does all the processing needed to build a specialised Nuance grammar from scratch. You will need to have defined at least the following config file parameters: \begin{itemize}

\item {\tt ebl\_\-corpus} The training corpus, which should consist of Prolog records of the form {\tt sent(...)}.

\item {\tt ebl\_\-operationality} The file defining the operationality criteria.

\item {\tt ebl\_\-nuance\_\-grammar} The output file which will contains the final Nuance grammar.

\end{itemize}

In many applications, you will also want the following:

\begin{itemize}

\item {\tt ebl\_\-include\_\-lex} A file of ``include-lex'' definitions, which specify lexical entries to be included directly.

\item {\tt ebl\_\-regulus\_\-component\_\-grammar} The specialised grammar that will be loaded by the {\tt EBL\_LOAD} command, if it is not the default one.

\item {\tt ebl\_\-ignore\_\-feats} Features to ignore when performing Regulus-to-Nuance compilation on the specialised grammar.

\end{itemize}