1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
|
<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="content-type"
content="text/html; charset=ISO-8859-1">
<title>SpamBayes</title>
</head>
<body>
<table width="100%" cellspacing="0" cellpadding="0">
<tbody>
<tr>
<td width="380"><img src="docs/images/sblogo.jpg" title=""
alt="Logo" style="width: 380px; height: 77px;"></td>
<!-- use background instead of styles to avoid mozilla bug 167262 -->
<td background="docs/images/span.jpg"> </td>
<td width="78"><img src="docs/images/python.jpg"
style="width: 78px; height: 77px;"></td>
</tr>
</tbody>
</table>
<h1>SpamBayes Server/Proxy Application</h1>
<strong>A spam filter based on statistical analysis of your personal mail.</strong>
<h2>Welcome to SpamBayes</h2>
<p><em>Note that if you use Microsoft Outlook, you should probably be using the
Outlook Addin, which was presented as an option at installation time.</em>
If you are using Outlook Express or almost any other mail program,
you are in the right place.</p>
<p>SpamBayes is an application that provides spam classification based on statistical
analysis of your personal mail. Unlike many other spam detection systems, SpamBayes
actually learns what you consider spam, and continually adapts as both your regular
email and spam patterns change.</p>
<p>This SpamBayes application sits between your mail server (generally at
your Internet Service Provider) and your mail program. Your mail program still believes
it is talking to a real mail server, but instead it is talking to
SpamBayes, which acts as a <span style="font-style: italic;">proxy</span>
between the two.</p>
<p>All your mail arrives as normal in your mail program, but, as it arrives,
SpamBayes adds a classification to the messages - either <b>ham</b> (good
mail), <b>spam</b> (bad mail), or <b>unsure</b> (messages that SpamBayes isn't
certain about). You can set your mail program up to automatically
filter messages into folders based on this classification - for example you
might move all <b>spam</b> messages to a folder called "Junk Mail", and all
<b>unsure</b> messages to a folder called "Mail to review".</p>
<p>In order to classify messages, you have to <i>train</i> SpamBayes. The
easiest method of doing this is via an interface that SpamBayes presents
to you via your web browser. SpamBayes will list all the messages that have
been received, and let you train them as <b>ham</b> or <b>spam</b>. You
can also configure SpamBayes via this interface in your web browser.</p>
<p>Please remember that this is free software; please be patient, and
note that there are plenty of things to improve. There are ways
you can help, even if you aren't a programmer - you could help with
this documentation, or <a
href="http://spambayes.org/donations.html">make a donation</a>, or
any number of other ways - please see the <a
href="http://spambayes.org">main project
page</a> for information. The list of <a
href="http://spambayes.org/faq.html">Frequently Asked Questions</a>
may also answer some of yours, especially considering the (current)
lack of documentation - you may wish to start in the
<a href="http://spambayes.sourceforge.net/faq.html#using-spambayes">
"Using SpamBayes" section</a>.</p>
<h3>Getting Started</h3>
<p><img src="docs/images/tray.jpg" style="float:right" alt="Tray menu" />
You should start the <span style="font-style: italic;">SpamBayes Tray
Icon</span> program, which was installed into your Start menu. This is the
recommended method of operating SpamBayes. If you right-click on this
icon you will be presented with a number of options to control SpamBayes
(see figure on the right), including an option to configure SpamBayes for
your system. Double-clicking on the icon will bring up the messages that
SpamBayes has for you to review.</p>
<p>For the most part, SpamBayes just silently works. When you do need
to interact with it, you do so via an interface presented in a web browser.
If you need help using the interface, click the "Help" button at the bottom
of any page. This also explains how you can contact the developers, if
you wish to do so.</p>
<h3>Using the Service</h3>
<p>There is a Windows service program installed, but it is not configured
by the installation program. If you configure this manually, you
can continue to use the Tray Icon program, as it will control the
service instead of running the server internally as it normally does.
If you don't know what a "Windows service program" is, don't worry about it;
you don't need to use this.</p>
<p>To install the service, perform the following steps:<br/>
<ul>
<li>Open a command prompt, and change to the <span
style="font-style: italic;">\Program Files\SpamBayes\bin</span>
directory.</li>
<li>Execute <span style="font-style: italic;">sb_service.exe
-install</span></li>
<li>Select <span style="font-style: italic;">Control
Panel->Administrative Tools->Services</span>, and locate <span
style="font-style: italic;">SpamBayes Service</span> in the list.</li>
<li>Change the properties of the service so it logs on with your user
account, rather than the builtin system account. This will ensure
that SpamBayes uses the same configuration and data files when running
as a service and when running as a normal program.</li>
<li>If desired, change the properties of the service to start at boot
time.</li>
<li>Start the service.</li>
<li>Start the <span style="font-style: italic;">SpamBayes Tray Icon</span>
program, and confirm the server is running. Configure and manage
SpamBayes normally.</li>
</ul>
<h3>Configuration</h3>
<p><img alt="Config page" style="float: right; margin: 1em" src="docs/images/config.jpg" />
There are two configuration steps that you need to take:</p>
<ol>
<li><strong>Setup SpamBayes:</strong> SpamBayes comes with default settings
that, for the most part, should be suitable for your system. There is
one setting that you will need to fill out, however, which is the address of
your mail server (at your ISP, most likely). Right-click on the tray icon
and choose <em>Configure...</em>; this will open up a browser window like the
one in the figure on the right. You can skip over most of the settings here,
apart from the top one (shown in the figure). You need to enter the address
of your mail server. If you don't know the address, you can find it in the
settings of your mail client.
<br /><br />
If you wish to, you can change other settings on this page - to the right
of each option is a description of what it does. There are also some
more advanced options available via a button at the end of the page
(note that these are not all of the SpamBayes options - see the
<a href="http://spambayes.org/faq.html">FAQ</a> for information about how
to access other options).
<br /><br />
Once you are satisfied that you have entered the correct options for your
setup, click the <em>Save</em> button at the end of the page. These
values will be saved in your personal configuration file and be used
each time SpamBayes is started up.
<br /><br />
<li><strong>Setup your mail client:</strong> You need to set your
mail client to receive mail from <em>localhost</em> rather than your
mail server.</li>
</ol>
<h3>Training</h3>
<p>When you first start using SpamBayes, all your mail will be classified as
'unsure' because SpamBayes doesn't have any preconceived ideas about what
looks like good or bad mail. As soon as you start training the classification
will improve, and by the time you've classified even 20 messages of each you'll
be seeing quite reasonable results.</p>
<p><img alt="Empty review page" style="float:right; margin: 1em" src="docs/images/review.jpg" />
SpamBayes saves a <strong>temporary copy</strong> of all incoming mail so that
classification can be independant of whatever mail client (Outlook Express, Eudora, ...)
you are using. You need to run through these messages and tell SpamBayes how to
handle mail like that in the future.</p>
<p>If you double-click the tray icon a browser window will open looking like the
figure on the right. For each message listed, you need to choose to either
discard (don't train on this message), defer (leave training on this message
until later), or train (as either good - <em>ham</em>), or bad - <em>spam</em>).
You do this by simply clicking in the circle in the appropriate column; if you
wish to change all the messages to the same action, you can simply click the
column heading.</p>
<p>You are presented with the subject and sender of each message, but, if this
isn't enough information for you to make a decision on the message, you can also
view the message text (this is the raw text, so you can't do any damage if the
message contains a virus or any other malignant data). To do this, simply click
on the subject of the message.</p>
<p>Once you have chosen the actions you wish to perform on all the displayed
messages, click the <em>Train</em> button at the end of the page. SpamBayes
will then update its database to reflect this data.</p>
<p>Note that the messages are split up into the classification that SpamBayes
would place the message with current training data (if this is correct, you might
choose to <em>Discard</em> the message, rather than train on it - see the
<a href="http://entrian.com/sbwiki">SpamBayes wiki</a> for discussion of training
techniques). You can also see the <em>Tokens</em> that the message contains
(the words in the message, plus some additional tokens that SpamBayes generates)
and the <em>Clues</em> that SpamBayes used in classifying the message (not all
tokens are used in classification).</p>
<p>So that the page isn't overwhelmingly long, messages waiting for review are
split by the day they arrived. You can use the <em>Previous Day</em> or
<em>Next Day</em> buttons at the top of the page to move between days. If mail
arrives while the review page is open the new messages will <strong>not</strong>
be automatically added to the displayed list; to view the new message, click
the <em>Refresh</em> button at the top of the page.</p>
<h2>Cost, Copyright and Licensing</h2>
<p>SpamBayes is copyright (C) 2002-4 Python Software Foundation; All Rights
Reserved.<br />
The Python Software Foundation (PSF) holds copyright on all material
in this project. You may use it under the terms of the PSF license;
see LICENSE.txt.</p>
<p>What all this means, in general terms, is that SpamBayes is completely
free to use, in pretty much any way you like. You're welcome to distribute
it, or even create your own software based on it. Enjoy!</p>
<h2><a name="Your_help_is_needed"></a>Your help is needed!</h2>
<p>This is free software. Please offer any help you are able
to. In particular, contributions to this documentation are
welcome! If you don't know where to start, the
<a href="http://spambayes.org/faq.html">FAQ</a> has
<a href="http://spambayes.org/faq.html#i-m-not-a-programmer-but-want-to-help-out-what-can-i-do">
details about how you can help</a>.
</p>
</body>
</html>
|
