Tag Archives: spam

Improved Filtering of SPAM and Phishing Email

ITS has begun enabling new, proactive anti-spam and anti-phishing email filters intended to improve the filtering of unsolicited and unwanted messages, by leveraging external reputation-based lists of email servers that have been flagged for sending spam and/or malicious email.

We’ve had the new filters in “Tag” mode since October 2014, so we’re confident that they are accurate and we’ll be closely monitoring incoming mail queues during the days after the change. Still, there is a possibility that someone trying to sending you email will have their message blocked, if their email servers have been flagged for sending spam and/or malicious email. The sender, in such cases, will receive an email advising them that their message could not be delivered.

If you encounter such as scenario, (i.e. a legitimate sender is trying to email you, but the message is being blocked by Middlebury’s email servers), or you have noticed legitimate messages recently having been incorrectly tagged as [SPAM?], please contact the ITS Help Desk at helpdesk@middlebury.edu and we will help you identify and resolve the issue.

General questions about this change may be directed to infosec@middlebury.edu.

Looking back at comment-spam in WordPress

In February 2012 we started noticing a large influx of new comment-spam coming into our sites.middlebury.edu WordPress system that the built-in anti-spam plugins weren’t able to handle. To combat this annoying plague we created a new plugin that instantly killed any comments trying to submit an “author URL” along with the “author name” and “comment text” now that the “author URL” field is hidden.

In the year and a half since this plugin has been in place across our blog network it has blocked an average of 40,000 spam comments every month.

+------+-------+--------------+
| year | month | spam blocked |
+------+-------+--------------+
| 2012 |     3 |       14,814 |
| 2012 |     4 |       19,956 |
| 2012 |     5 |       18,225 |
| 2012 |     6 |       15,937 |
| 2012 |     7 |       29,232 |
| 2012 |     8 |       24,073 |
| 2012 |     9 |       25,973 |
| 2012 |    10 |       42,514 |
| 2012 |    11 |       49,265 |
| 2012 |    12 |      106,128 |
| 2013 |     1 |      103,850 |
| 2013 |     2 |       72,944 |
| 2013 |     3 |       38,336 |
| 2013 |     4 |       35,125 |
| 2013 |     5 |       32,975 |
| 2013 |     6 |       35,011 |
| 2013 |     7 |       28,218 |
+------+-------+--------------+

While some spam is bound to get past any automated filtering, we hope that these efforts have alleviated most of the hassle of dealing with spam comments in WordPress.

Reduced comment spam in blogs

During the past few months we have been seeing an increased amount of comment spam coming into WordPress (sites.middlebury.edu) that follows a distinctive pattern: the comment text is useless, but unoffensive and contains no links itself, while the Comment Author Website field contains the URL of a commercial site. Because the comment text doesn’t contain any links, the comment doesn’t get picked up by WordPress’s existing spam filters and until now would be held for moderation. Continue reading

Does tagging content make it easier to find with search? No.

I’ve received this question from several people now. Below are two videos from Matt Cutts who works on Google’s Webspam team explaining how tagging content mostly does not affect their search results. This also means that tagging largely will not affect how results appear on Middlebury’s site, since we use Google to provide our search results.

Tags

Tag Clouds

This does not mean that you shouldn’t tag content at all. Tags can still be useful for humans who want to find other posts and pages on a topic. However, if you want your page to be easier to find, your time is better invested in making sure that the content is well written, structured and relevant to a particular topic.

PCI and Blocked Email Messages

The Payment Card Industry Data Security Standard (PCI DSS v2.0) is a standard that has been accepted by all major credit card companies and most credit providers. It is a standard that we must abide by if we are to accept credit cards as a form of payment. PCI DSS is broken into 12 requirements; each focusing on a different domain of security.

While PCI DSS is not an actual law, it is a standard enforced by the credit card industry, and the banks have stated and upheld the policy that they will no longer accept business from non-PCI compliant merchants. The government has used the PCI DSS as a yardstick by which they have measured such regulations as Gram-Leach-Bliley, Sarbanes-Oxley, and most recently the drafting of the Data Accountability and Trust Act.

We employ a device called a Barracuda here at Middlebury which helps us prevent SPAM from flooding our email system. Just shy of a year ago this system was updated to enable it to filter on cardholder information. By default this feature was turned on. We have left this enabled and have begun reporting on these blocked messages and alerting the senders of outbound messages.  The Barracuda is intended to serve both as a SPAM filter and a compliance tool.