Welcome to pab's potpourri!

This is an experimental, informal blog for learning about blogging, blog development, and blog-related professional development activities.

pab's potpourri

↑ Grab this Headline Animator

Saturday, September 29, 2007

Gmail spam filters add warning: a follow-up

This is a follow-up on Gmail spam filters add warning (January 4, 2007). It suggests shortcomings both in legitimate mail senders' message subject heading practices, and in Gmail's analysis of innocuously entitled messages.

During an occasional foray into the spam folder this weekend, to grapple with the 1800+ messages that had accumulated since my previous foray a few weeks ago, I discovered only three false positives (approximately 0.17%):

  1. Shots, Shows, And Shirts, from Snap Shots
  2. Edublogs and EduBling!, from Edublogs
  3. Sept sessions and Nov online event, from Cathy Baxter
The first two messages may have gotten snagged because of their snazzy subject headings, even though the senders could be well-known, and their mail servers verifiable.

However, the third one, which was the first that I had detected in rapid glances at twenty-five senders per page, was a noodle twister. I had recognized the sender on first glance, and the subject was innocuous.

Nevertheless, a gmail-generated warning superimposed on body of the message suggests that the links in the message itself were the problem:

Warning: This message may not be from whom it claims to be. Beware of following any links in it or of providing the sender with any personal information. Learn more
([within] personal correspondance, September 24, 2007)

Indeed, inside were truncated, somewhat destination-opaque links through tinyuurl.com to Eluminate Live! sessions (closed Sep 26, 2007):
Although those links were in a message from a recognized sender, and in context of a legitimate, far less destination-opaque link:
Gmail spam filters apparently caught those tiny URLs!

2 comments:

  1. In a similar foray into my Gmail spam folder on June 7, 2008, I waded through over 2000 messages that had accumulated since May 18. On that foray I discovered only one false positive - less than 0.05%. Fantastic!

    ReplyDelete
  2. It is about nine months since my last Gmail spam assessment (previous comment, 2008.06.07). Nevertheless, it is a pleasure to announce that although I failed to note the number of messages in the spam folder before I reviewed them and purged it today, the false positive ratio to however many messages remained since my last perusal (no date) was 0.0%.

    ReplyDelete

Local time is:

(JST)