mailcross.1

dbacl是一个通用目的的digramic贝叶斯文本分类器。它可以学习你提供的文本

\" t
.TH MAILCROSS 1 "Bayesian Text Classification Tools" "Version 1.3" ""
.SH NAME
mailcross \- a cross-validation tester for use with dbacl.
.SH SYNOPSIS
.HP
.B mailcross 
.I command 
[
.I command_arguments 
]
.SH DESCRIPTION
.PP
.B mailcross
automates the task of cross-validating email filtering and classification
programs such as 
.BR dbacl (1).
Given a set of categorized documents, mailcross initiates test runs 
to estimate the classification errors and thereby permit fine tuning 
of the parameters of the classifier. 
.PP
Cross-validation is a method which is widely used to compare the quality 
of classification and learning algorithms, and therefore permits rudimentary
comparisons between those classifiers which make use of  
.BR dbacl (1)
and 
.BR bayesol (1),
and other competing classifiers. 
.PP
The mechanics of cross-validation are simple: A set of pre-classified 
email messages is first split into a number of roughly equal-sized subsets.
For each subset, the filter (by default, 
.BR dbacl (1)) 
is used to classify each message within this subset, based upon having learned 
the categories from the remaining subsets. The resulting classification errors are then averaged over all subsets.
.PP
.B mailcross
uses the environment variables MAILCROSS_LEARNER and MAILCROSS_FILTER when
executing, which permits the cross-validation of arbitrary filters, provided
these satisfy the compatibility conditions stated in the  
ENVIRONMENT section below.
.PP
During preparation, 
.B mailcross
builds a subdirectory named mailcross.d in the current working directory. 
All needed calculations are performed inside this subdirectory.
.SH EXIT STATUS
.B mailcross
returns 1 on success, 0 if a problem occurred.
.SH COMMANDS
.PP
.IP "\fBprepare\fR \fIsize\fR"
Prepares a subdirectory named mailcross.d in the current working directory, and
populates it with empty subdirectories for exactly 
.I size
subsets.
.IP "\fBadd\fR \fIcategory\fR [FILE]..."
Takes a set of emails from either FILE if specified, or STDIN, and 
associates them with 
.IR category .
All emails are distributed randomly into the subdirectories of mailcross.d for later use. For each
.IR category , 
this command can be repeated several times, 
but should be executed at least once.
.IP "\fBclean\fR"
Deletes the directory mailcross.d and all its contents.
.IP "\fBlearn\fR"
For every previously built subset of email messages, pre-learns all 
the categories based on the contents of all the subsets except this one.
The 
.I command_arguments
are passed to MAILCROSS_LEARNER.
.IP "\fBrun\fR"
For every previously built subset of email messages, performs the classification
based upon the pre-learned categories associated with this subset.
The 
.I command_arguments
are passed to MAILCROSS_FILTER.
.IP "\fBsummarize\fR"
Prints statistics for the latest cross-validation run.
.SH USAGE
.PP
The normal usage pattern is the following: first, you should separate your email
collection into several categories (manually or otherwise). Each category should
be associated with one or more folders, but each folder should not contain 
more than one category. Next, you should decide how many subsets to use, say 10. 
Note that too many subsets will slow down the calculations rapidly. Now you can type
.HP
.na
% mailcross prepare 10
.ad
.PP
Next, for every category, you must add every folder associated with this
category. Suppose you have three categories named 
.IR spam , 
.IR work , 
and 
.IR play ,
which are associated with the 
.SM BSD 
mbox files 
.IR spam.mbox , 
.IR work.mbox , 
and 
.IR play.mbox 
respectively. You would type
.PP
.na
% mailcross add spam spam.mbox
.br
% mailcross add work work.mbox
.br
% mailcross add play play.mbox
.ad
.PP
You can now perform the cross-validation. Note that the learning stage can take
some time. Typically, you would learn once, and run several times with 
varying risk parameters in case you use 
.BR bayesol (1).
.PP
.na
% mailcross learn
.br
% mailcross run
.br
% mailcross summarize
.ad
.PP
Once you are all done cross validating, you can delete the working files, log
files etc. by typing
.PP
.na
% mailcross clean
.ad
.SH ENVIRONMENT
.PP
Right after loading, 
.B mailcross 
reads the hidden file .mailcrossrc in the $HOME directory, if it exists, so
this would be a good place to define custom values for environment variables.
.IP MAILCROSS_FILTER
This variable contains a shell command to be executed repeatedly
during the running stage.
The command should accept an email message on STDIN and output a 
resulting category name. It should also accept a list of category file names
on the command line, each preceded by the string "-c". 
If undefined, 
.B mailcross
uses the default value
MAILCROSS_FILTER="dbacl -T email -v".
.IP MAILCROSS_LEARNER
This variable contains a shell command to be executed repeatedly during the
learning stage. The command should accept a 
.SM BSD 
mbox type stream of emails
on STDIN 
for learning, and the file name of the category on the command line. 
If undefined, 
.B mailcross
uses the default value
MAILCROSS_LEARNER="dbacl -T email -l".
.SH NOTES
.PP
The subdirectory mailcross.d can grow quite large. It 
contains a full copy of the training corpora, as well as learning files for 
.I size 
times all the added categories, and various log files. 
.SH WARNING
.PP
Cross-validation is a widely used, but ad-hoc statistical procedure, completely
unrelated to Bayesian theory, and subject to controversy. 
Use this at your own risk.
.SH SOURCE
.PP
The source code for the latest version of this program is available at 
http://www.lbreyer.com/gpl.html
.SH AUTHOR
.PP
Laird A. Breyer <[e-mail]>
.SH SEE ALSO
.PP
.BR bayesol (1)
.BR dbacl (1), 
.BR mailinspect (1),
.BR regex (7)