Currently, the official CWB documentation consists of two extended tutorials, which introduce the CQP query language, interactive use of the CQP query processor, and the corpus encoding procedure. We intend to add detailed specifications and reference manuals at a later time.
This tutorial is essential reading for all CWB users, and serves as a comprehensive user manual for the CQP query processor (as an interactive command-line program). In order to try the examples in the tutorial, either download the pre-encoded sample corpora or use the online CQP demos.
This guide is intended for corpus administrators and all other users who want to index their own corpora or extract frequency information from a CWB corpus. It is highly recommended to install the CWB/Perl packages first, which include various tools to simplify and optimise the corpus encoding procedure.
Please note that the versions linked above document CWB version 3.0 only. More recent versions are available in the Subversion repository.
As of CQPweb version 3.1.0, there is a system administrator's manual, available within the CQPweb code tree itself (under doc) as well as from this website. Currently it is available as a PDF, an HTML verison is coming soon. The manual explains a number of topics, including the installation process for CQPweb.
Please note that this is not targeted at users of CQPweb, but rather at superusers who are actually running a server (or running CQPweb on a standalone machine). It assumes a basic knowledge of administration for Unix-like systems in general.
CWB has a YouTube channel! Videos posted here are directed, in particular, at the graphical web interfaces (CQPweb and BNCweb, so far) and are more beginner-friendly than some of the manuals above.
For further help, consult the FAQ and the CWB wiki. You should also join the friendly and helpful community on the CWB mailing list. Don't hesitate to ask questions about installation problems, general usage and CQP query tips & tricks there.