Since I’ve redone the design on my blog, I figured that it would be nice to enable comments again. I’ve stumbled over Steve Kemp’s blogpost “When the light is green, the trap is clean.” which is about a XML-RPC based antispam service for blogs and forums. I followed the link to blogspam.net and read a bit about the API. It seemed pretty straightforward and as an adition, it was open source.
There was no java implementation, so I made one, based on the apache XML-RPC client (project site) and it only implements the basics. The sources are up for grabs here.
Usage:
To test if a comment is spam or not, you should use the function isSpam.
Example:
HashMap options = new HashMap();
options.put("comment","Ten animals I slam in a net");
options.put("ip","91.193.130.226");
String result = (new BlogSpamTest()).isSpam(options);
Now the string result contains the response sent from the blogspam.net server.
To train the bayasian filtering system, use the train method which take to parameters:
comment – A string containing the comment to be trained upon.
spam – A boolean. true if the comment is spam, false otherwise.
Example:
/*
* Tell the server that following is spam, do this:
* Buy cheap viagra http://viagrashop.nl
*/
(new BlogSpamTest()).train("Buy cheap viagra http://viagrashop.nl",true);
/*
* On the other hand, if the server classified the sentence as spam:
* That is one nice car.
*
* and you don't think that it is, do this:
*/
(new BlogSpamTest()).train("That is one nice car.",false);
If you use it, improve it or have any questions, please let me know.
UPDATE:
All methods have been made static.
I have added the getStats method, which takes one string as a argument, the string being the hostname to get stats for.
Copyright © 2010 Jannich Brendle's blog All rights reserved. Amazing Grace theme by Vladimir Prelovac.
Using blogspam.net for antispam
Since I’ve redone the design on my blog, I figured that it would be nice to enable comments again. I’ve stumbled over Steve Kemp’s blogpost “When the light is green, the trap is clean.” which is about a XML-RPC based antispam service for blogs and forums. I followed the link to blogspam.net and read a bit about the API. It seemed pretty straightforward and as an adition, it was open source.
There was no java implementation, so I made one, based on the apache XML-RPC client (project site) and it only implements the basics. The sources are up for grabs here.
Usage:
To test if a comment is spam or not, you should use the function isSpam.
Example:
HashMap options = new HashMap();options.put("comment","Ten animals I slam in a net");
options.put("ip","91.193.130.226");
String result = (new BlogSpamTest()).isSpam(options);
Now the string result contains the response sent from the blogspam.net server.
To train the bayasian filtering system, use the train method which take to parameters:
Example:
/** Tell the server that following is spam, do this:
* Buy cheap viagra http://viagrashop.nl
*/
(new BlogSpamTest()).train("Buy cheap viagra http://viagrashop.nl",true);/*
* On the other hand, if the server classified the sentence as spam:
* That is one nice car.
*
* and you don't think that it is, do this:
*/
(new BlogSpamTest()).train("That is one nice car.",false);If you use it, improve it or have any questions, please let me know.
UPDATE:
All methods have been made static.
I have added the getStats method, which takes one string as a argument, the string being the hostname to get stats for.
Example:
HashMap stats = BlogSpamTest.getStats("http://blog.bredsaal.dk");System.out.println("Recieved " + stats.get("ok") + "good comments and " + stats.get("spam") + " spam comments.");
Related posts:
Tags: antispam, commentspam, java