<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
		>
<channel>
	<title>Comments on: A New Battle in the War on Spam</title>
	<atom:link href="http://www.robhyndman.com/2006/10/12/a-new-battle-in-the-war-on-spam/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.robhyndman.com/2006/10/12/a-new-battle-in-the-war-on-spam/</link>
	<description>any technology distinguishable from magic is not sufficiently advanced</description>
	<lastBuildDate>Thu, 12 Apr 2012 14:32:35 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.3.2</generator>
	<item>
		<title>By: Ben Lucier</title>
		<link>http://www.robhyndman.com/2006/10/12/a-new-battle-in-the-war-on-spam/comment-page-1/#comment-18661</link>
		<dc:creator>Ben Lucier</dc:creator>
		<pubDate>Fri, 13 Oct 2006 13:54:36 +0000</pubDate>
		<guid isPermaLink="false">http://www.robhyndman.com/2006/10/12/a-new-battle-in-the-war-on-spam/#comment-18661</guid>
		<description>I&#039;m confident that SPAM software will catch up, but it will become more and more important to combat spam at the service provider level.  

Tyler&#039;s piece is accurate in that more spam is being delivered through digital images.  Without getting into too much detail (maybe I&#039;ll post my own entry soon), service providers are beginning to implement OCR technology to &quot;read&quot; the images and convert the image into written text that can then be spam filtered.  

While this sounds CPU intensive, the systems are pretty efficient: Since the spammers use the same image, sent to millions of different users, after the image is OCR&#039;ed, a unique hash is generated and stored to immediately identify other matching images, without the cpu intensive OCR method.  

This is why it&#039;s important to filter SPAM at the service provider level - stop it at the source.</description>
		<content:encoded><![CDATA[<p>I&#8217;m confident that SPAM software will catch up, but it will become more and more important to combat spam at the service provider level.  </p>
<p>Tyler&#8217;s piece is accurate in that more spam is being delivered through digital images.  Without getting into too much detail (maybe I&#8217;ll post my own entry soon), service providers are beginning to implement OCR technology to &#8220;read&#8221; the images and convert the image into written text that can then be spam filtered.  </p>
<p>While this sounds CPU intensive, the systems are pretty efficient: Since the spammers use the same image, sent to millions of different users, after the image is OCR&#8217;ed, a unique hash is generated and stored to immediately identify other matching images, without the cpu intensive OCR method.  </p>
<p>This is why it&#8217;s important to filter SPAM at the service provider level &#8211; stop it at the source.</p>
]]></content:encoded>
	</item>
</channel>
</rss>

