David Jackson's  Instablog

David Jackson
Send Message
I'm founder and CEO of Seeking Alpha. I started Seeking Alpha after working for five years as a technology research analyst for Morgan Stanley in New York. Seeking Alpha is now the dominant crowdsourced equity research platform, with thousands of contributors and millions of monthly visitors,... More
My company:
Seeking alpha
My blog:
A Founder's Notebook
  • How to Stop Spam 5 comments
    Jun 24, 2009 4:32 AM

    You know spam when you see it: it's lousy for the reader. I'm sure that Seeking Alpha isn't the only website attacked by spammers. In our case, we see two types of spammers: people trying aggressively to promote their financial businesses (at least they're relevant to our readers), and those publishing blogs and other content that have no relevance whatsoever to our readers.

    In our efforts to combat spam, we thought about the fundamental principles of how to win the war against spammers. We came up with these principles:

    1. It must be easy for us to find spam content.
    2. It must be easy for us to delete spam content.
    3. It must be significantly less effort for us to find and delete spam than for the spammer to post the spam to our website.
    4. Our anti-spam tools must not create a high barrier for our readers to contribute genuine content to Seeking Alpha.

    We rely on our readers to help us identify spam -- we have "report abuse" links below every comment, for example. And we've built dashboards that allow our editors to delete a spammer's identity and content in one click, while barring that email address from future use. And only StockTalks written by users that have uploaded a picture can appear on our home page.

    What is your experience with spam, and what recommendations do you have for how to deal with it?

Back To David Jackson's Instablog HomePage »

Instablogs are blogs which are instantly set up and networked within the Seeking Alpha community. Instablog posts are not selected, edited or screened by Seeking Alpha editors, in contrast to contributors' articles.

Comments (5)
Track new comments
  • H. T. Love
    , contributor
    Comments (19539) | Send Message
    It's difficult to automate this 100%. For e-mail submisions there's lots of free software packages that assign weightings based on certain attributes that are user-modifiable and tunable.


    Maybe some of these could be adapted to run over whatever Content Management System (CMS) you use?


    Assuming that poses to many technical hurdles, I think a reasonble-effort manual/mechanical aided system could do the job.


    If a database or file of previously seen and "approved" pseudonyms, or whatever else you allow, could be created (a mechanical "dump" of all existing IDs would do for the initial load) and maintained, then whenever a post appeared it could be checked against this DB. If it was not there, it would be flagged and posted to a person for review. Until review was completed, the post would be allowed just as now, but would continue to be flagged for review until a review was completed.


    If the reviewer says it looks ok, the flag would be changed to ok. If not, the flag would be changed to "spammer".


    If a flag is now ok, any future posts would only trigger review via the current method - user reports. If flagged for review, continue to remind the reviewer. If flagged for spam, future posts are rejected. A mesage to the poster is suggested (maybe a "Please e-mail xxxx if you believe an error ...") so that normal types of mistakes can be corrected and remove the "spammer" flag.


    Whether this remains in the same DB or a "spammer" DB is really a performance consideration. If a separate DB, then it would be checked before the new/approved DB.


    Normal maintenance support, add, change, delete, addresses and status would be needed.


    Because new e-mail, web sites, etc. are so easily established these days, another possible "trigger for review" might be based on previously seen IP addresses. The weakness here is DHCP makes IPs not static, so that's not really reliable. But there are "groupings" that could be identified as more likely to be spammers and trigger a review flag whenever a new poster with an IP in that range (block) is seen. GeoIP-enabled IP, common now, might also be usable for that.


    But I do think that will take more work and may not justify the effort needed.


    26 Jul 2009, 01:55 PM Reply Like
  • H. T. Love
    , contributor
    Comments (19539) | Send Message
    On Jul 26 04:24 PM Freya wrote:


    > HTL: why bother? just eliminate them.


    I thought he was looking for suggestions on how to more effectively identify and eliminate them. As example the "iamned" spammer, keeps coming back by using different names. This has a cost to us and to SA. If they had a mechanized assistant, it would reduce their cost and reduce the impact on us.


    Oh well, it's not the first time I've misunderstood the intent of a request.




    > Spam is Spam. Those commentators who are not Spammers will adjust
    > their contents. Those that Spam will see their efforts deleted. and
    > if these efforts continue, then the spammers themselves, can be deleted.
    > One "liners" do not give any Article the Respect it deserves. I say
    > "Respect" because whether you agree with the Gist of the Article
    > or Not, the Author deserves to know why.
    26 Jul 2009, 07:46 PM Reply Like
  • one eye
    , contributor
    Comments (645) | Send Message
    It should not be just limited to Commentators. Authors also contribute by Spamming their Websites in every "comment" they make.


    I think "iamned" is Cetin reborn.


    James Quinn and Peter Cooper are just two others. Mad Hedge is worse. His posts are really large but they usually have one thing in common: the same comment in multiple Articles like Clockwork. He has a few variations but if you go back a month, 2 months, 3 months, you will find that on the whole they are the same: "Ouch" was my favorite but an "evening with Yellen and Rubin" are up there too.
    27 Jul 2009, 01:42 AM Reply Like
  • thiazole
    , contributor
    Comments (2236) | Send Message
    I got frustrated with the IamNed spammer and found every spam post he made under every alias that day and reported abuse on all of them - it was very effective until it got me banned for a weekend (I appealed and got unbanned a couple days later, but the damage to me was done as I missed commenting on what I believed were some very important articles). Apparently someone thought I was abusing the "report abuse" button. What we really need is a "report spam" button, since it isn't always that obvious that a post is "abuse" when it is clearly just spam. It would also alert people to report spam, whereas I don't believe many are reporting it right now. In any event, I am now afraid to report abuse much any more, although I still report a few IamNed ads every day.


    On Jul 26 08:53 AM Freya wrote:


    > Besides, doing so could get you deleted if the Wrong Staffer sees
    > you doing it.
    6 Aug 2009, 03:37 PM Reply Like
  • thiazole
    , contributor
    Comments (2236) | Send Message
    BTW, looking at this thread, it is obvious that someone is giving Freya here a thumbs down on every post he/she makes. Perhaps that should be investigated...
    6 Aug 2009, 03:40 PM Reply Like
Full index of posts »
Latest Followers


More »

Latest Comments

Instablogs are Seeking Alpha's free blogging platform customized for finance, with instant set up and exposure to millions of readers interested in the financial markets. Publish your own instablog in minutes.