POPFile - Automatic Email Classification / News: Recent posts

There's a great article (http://www.usatoday.com/tech/columnist/andrewkantor/2004-03-26-kantor_x.htm) in USA Today about Open Source alternatives to closed source software. Amongst many programs they say the following about POPFile:

"PC Magazine recently ran a comparison of anti-spam products. Unfortunately, it left out a lot of them, namely the myriad of open-source solutions (perhaps because open-source programmers aren't likely to buy advertising). One of the best anti-spam products in fact is POPFile, a Bayesian spam filter that can not only sort out your spam from your real mail, but even learn to sort that real mail into subject areas. (Unlike other spam filters, POPFile doesn't just separate "good" and "spam" messages. You can have any number of mail "buckets," although dedicating one for spam is the most popular use.)"... read more

Changes in v0.21.1

This release was prompted by a number of users who saw very
long corpus conversion times when upgrading to v0.21.0. This
has been fixed along with a small number of others things:

1. Updated Arabic translation
2. Updated installer to handle corpus upgrade problems better
3. Logger always uses system specific line endings
4. A side effect of fixing the upgrade slowness is that
reclassification will be faster.
5. Made FAQ, ID and (stealth mode) localizable
6. Make old bucket detection more robust
7. Fix problem on NextStep with directory creation
8. Fix slurp_ so that it doesn't break CRLF on slow links.... read more

Welcome to POPFile v0.21.0

This version consists of a major update to v0.20.1 with many improvements
and bug fixes:

1. Multi-user phase 1

POPFile now recognizes two environment variables (POPFILE_ROOT and
POPFILE_USER) which can be used to customize the location of POPFile
and the location of its per-user data. POPFile *no longer* has to be
run with the current working directory set to where popfile.pl is
installed. ... read more

I'll be "going" to the MIT Spam Conference 2004 on January 16. I say "going" because in fact I wont be there in person, but will be presenting a talk entitled "How to beat an Adaptive Spam Filter" by video.

I'll make a copy of the talk available on the web once it's been shown.

John.

POPFile has been selected to alt.comp.freeware's Pricelessware list for the second year running:

http://www.pricelessware.org/2004/PL2004INTERNET.htm

John.

POPFile's started getting mentioned in some books, here's the scoop:

1. The Healthy PC: Preventive Care and Home Remedies for Your Computer, by Carey Holzman (http://www.careyholzman.com).

Some time ago the author told me that he was going to write about POPFile in his upcoming book, well that book's here and http://www.amazon.com/exec/obidos/ASIN/0072229233/extravalent-20 from Amazon.

Here's how the author describes it:... read more

To assist in the development of better POPFile documentation there's a new place to go... the POPFile Documentation Project:

http://popfile.sourceforge.net/cgi-bin/wiki.pl

Here you'll find documents contributed by me and by some very energetic POPFile users that cover many aspects of POPFile use and development. You too can contribute freely through the web site.

John.

INTRODUCTION

v0.20.1 is a minor update on v0.20.0 to address the performance problems that some
people were seeing with the introduction of the BerkeleyDB for storage of the corpus.
v0.20.1 sets the BerkeleyDB cache size (i.e. the amount of the corpus that is kept
in memory) to 64KB and it can be controlled through the bayes_db_cache_size
parameter if needed. The default should work for everyone, you can set this to a
higher value to keep more of the database in memory at the cost of using more
memory.... read more

I've put my Usenix LISA talk from yesterday up on the web; it's in evil PowerPoint Show format because my PDF converter seems to be dead this morning!

http://popfile.sourceforge.net/usenix2003.pps (600Kb)

John.

POPFile is an email classification tool with a Naive Bayes classifier, a POP3 proxy and a web interface. It runs on most platforms and with most email clients. v0.20.0 is a major update to POPFile with the focus being on performance. In addition POPFile makes another leap forward in support for non-English languages with many new UI translations (including our first right to left language) and full support for parsing of Japanese and Korean.... read more

I'm presenting and sitting on a panel at the LISA conference organized by Usenix in San Diego on October 30. Any POPFile users are welcome to stop by and say Hi either after the talk or during one of the "meet the speakers" sessions that day.

http://www.usenix.org/events/lisa03/
http://www.usenix.org/events/lisa03/tech/techthurs.html

Adaptive Filtering: One Year On
John Graham-Cumming, ActiveState ... read more

POPFile is one year old today.

Back on September 22, 2002 I released "POPFile beta 1.5" which consisted of a single Perl script plus a library from CMU for doing text classification. POPFile only worked on Windows.

The brief history of POPFile can be found here: https://sourceforge.net/docman/display_doc.php?docid=17906&group_id=63137

John.

John.

One of POPFile's dedicated users has launched a Japanese-only site for Japanese users of POPFile on the Japan-based Sourceforge.jp site.

It's here: http://popfile.sourceforge.jp/

Thanks, Junya Ishihara it looks great! (Hope it says nice things in Japanese :-)

John.

Read all about it here: http://www.webattack.com/get/popfile.html

I'll be appearing on Rich Levin's show PC Talk on WPEN 950 AM radio in Philadelphia. Info is here: http://www.rblevin.net/PCT.htm. In addition the show is web cast so you can listen from anywhere, get those details here: http://www.rblevin.net/Webcasts.htm

The show runs from 2000 to 2100 (Eastern Time) on August 3, 2003.

If you want to call in and bug me about something the number is 800-876-WPEN.... read more

INTRODUCTION

Today's release consists of two maintenance releases to v0.18.2 and
v0.19.0.

A special thank you goes out to Sam Schinke who continues to contribute
time and energy to POPFile.

I recommend v0.19.1 because of the many improvements, but v0.18.3 is the
STABLE release for those who want rock solid reliability and want others to
do the bug hunting on the new v0.19.x line :-)

ESSENTIAL READING IF YOU ARE UPGRADING TO v0.19.1 from v0.18.x... read more

The Courier Email client (http://www.rosecitysoftware.com/courier/) has code in the latest version to automatically find the X-POPFile-Link decode the link and add a "Reclassify..." menu item.

Very nice integration of POPFile. Thanks.

John.

I'll be appearing on "Special Report with Brit Hume" on Fox News Channel at 6pm Eastern (and repeated at 12am) talking about spam.

http://www.foxnews.com/specialreport/index.html

John.

INTRODUCTION

For the first time in POPFile history we are releasing two versions at the
same time. Please take a moment to decide which version is most appropriate
for you.

v0.18.2 is a minor bug fix update to the last v0.18.1 made in February and
contains no new features. It is intended for conservative users who are
happy with POPFile as it stands and would just like bug fixes and nothing
else.

v0.19.0 is a major update to v0.18.1 and include everything that is fixed
in v0.18.2 plus much more. Full details are below.... read more

If you are interested in promoting POPFile on your web site there are banners and buttons here: http://popfile.sourceforge.net/banners.html.

John.

Folks,

Meet Otto the POPFile octopus. You can see Otto by visiting http://popfile.sourceforge.net.

You'll find links to the standard POPFile manual, a page with buttons you can use to advertize POPFile and the new POPFile store.

John.

Read all about it.

See an odd picture of me.

http://sourceforge.net/potm/potm-2003-05.php

John.

Listen here: http://media.cmpnet.com:8081/ramgen/twtoday_media/nc/itminute/21apr03/21apr03.smil

John.

If you buy the May issue of Linux Magazine you'll see a cover story called "Inside the POPFile Project" which is a three page interview with me, Stan and Sam.

It's well worth the cover price :-)

John.

Folks,

Some of you may have seen press reports and a press release (see below for links) about POPFile and ActiveState. ActiveState asked me to be a consultant for them on their Antispam Task Force and I said "ok!'

I'll be using my experience on the spam side of POPFile to help improve ActiveState's products. This has no effect on the status of POPFile and there are no plans to commercialize POPFile, make it closed source, or any silliness like that.... read more