News

David Sifry has posted his latest State of the Blogosphere posting. Here is what he has to say about splogs:

There has been an increase in the overall noise level in the blogosphere, most notably in the number of spam and fake pings that are sent - what I call “spings”. These spam pings are fake or bogus notifications that a blog has been updated; in some cases, these spings can amount to a denial-of-service attack, and can sometimes account for as much as 60% of the total pings Technorati receives. However, we’ve built a sophisticated system that mitigates the spings, and helps to keep spam blogs out of our indexes. Beyond that, about 9% of new blogs are spam or machine generated, or are attempts to create link farms or click fraud. Technorati continues to take an ecosystem approach to solving this problem, working closely with other players like Amazon, AOL, Ask Jeeves, Drupal, Google, MSN, Six Apart, Tucows, Wordpress and Yahoo, and there will be another Web 2.0 Spam Squashing Summit this spring, building on the success of the previous two summits.

I suspect they are under counting the number of splogs out there. I have a couple of Technorati search feeds and see large amounts of splogs in the continually.

A2B has been reporting success in identifying and block splogs based on IP addresses. They have significantly reduced the load on their server by actively block sploggers. This seems like it is a very effective strategy for them. They have identified 442 servers that are dedicated to sending splogs.

Frank and Jeff, who started SplogReporter, are looking for someone to help out with their service.

Recently, Jeff and I have become extremely busy with our personal lives yet we continue to receive praise, questions and requests via our SR feedback page just about everyday. To be quite honest with you we cannot keep up with all of the feedback.

In realizing our short-handedness, we would be willing to bring on some help in the spirit of the movement to continue to move the product forward in a positive manner by developing out some of the requested features that we have received as feedback.

If you think you may be able to help out, let them know.

Randy Morin has posted a Splogger Challenge to www.301powered.com. This site was copying the full text of Randy’s posts along with other feeds. The owner of the site supplied a response, claiming the site was not splog, but a useful aggregation of feeds.

The two posts have generated a lot of comments on Randy’s site, debating whether or not the site is splog. I have to agree with Randy on this one for calling the creator of the site out. The site is re-posting full text of entire feeds without permission, clearly making it a copyright violation. Whether or not to call the site “splog” does not seem as clear cut. The site is different from typical splog which is usually not useful to anyone at all, and is created mostly to get hits from search engines. So I would say this is borderline splog.

Matt Cutts is asking for feedback about what Google should do about web spam in 2006.

Okay, I’m most interested in hearing what people think Google should tackle in webspam. Based on your experiences, close your eyes and think about what area(s) you wish Google would work on. You probably want to think about it for a while without viewing other people’s comments, and I’m not going to mention any specific area that would bias you; I want people to independently consider what they think Google should work on to decrease webspam in the next six months to a year.

I won’t bias you with my thoughts on it either, but you probably have a good idea of what I think the biggest problem is

A few more high profile bloggers have been complaining about splogs stealing content. Here’s John Battelle

We need to address this….

And Om Malik:

Anyway I think the problem of splogs and scraping can be fixed if Google steps’ up to the plate.

A new study shows that 75% of pings are generated from splogs. The report analysis pings from weblogs.com over a four week period.

we used our work on splog detection to detect splogs (and hence spings) among the english blogs. Our detection mechanism is close to 90% accurate. As shown in the charts below pings from blogs average around 8K per hour and those from splogs average around 25K.

This is the most comprehensive study of splogs I have seen, and the results are not that surprising to me. This shows that splogs are clearly out of control.

They are also publishing live statistics of blog languages and blogs vs. splogs at http://memeta.umbc.edu/

Found via: The Blog Herald

Mark Cuban has another blog entry dealing with splogs. This time he is responding to a complaint by Steve Rubel that his content is being stolen. Mark’s answer to the problem is to have Blog search engines not index splogs that steal content; only the original entry would be indexed.

This is only half the solution, though. It’s not just blog search engines that find blog content, regular search engines do too. It’s a little harder for regular search engines to filter out duplicate blog entries because they have no knowledge of individual entries, only pages, and its harder for them to know the originator of content if the content is stolen quickly.

While I was gone (because of a crashed hard drive, vacation and business travel, all in a row), Plagiarism Today posted an excellent article.

It pretty much nails all the issues that are relevant in the splog war today, and as often happens in articles about splog, puts Google right in the middle of things.

Google, however, has been a mixed ally in this matter. Though they took token steps to prevent splogs from being created on their service, those steps have proved inadequate and no further plans have been made.

One of the solutions suggested in the article, which makes sense to me, is cutting back on the automation in the blog world.

Perhaps it’s time that we removed some of the automation from the blogging world. Perhaps pings, trackbacks and comments should come with some form of authorization. Even a simple Hashcash setup, a system by which a user has to take a few seconds to complete a simple task before finishing an action, could go a long way to fix the problem.

It can be a hassle for legitimate bloggers, as was seen when Blogger introduced captchas for posting, it probably is time to start taking more serious steps.

The Guardian has published a splog article today. The article is fairly well written and gets seems to get things right. Again, most of the focus is on Google because they provide an infrastructure to create splogs with Blogger, and a way to earn money with AdSense.

Here is a quote from Jason Goldman, product manager for Blogger:

“Spam blogs cost Google money both on the hosting and infrastructure side for Blogger as well on the AdSense side with spam prevention,” says Goldman. “We take our obligation to our AdSense advertisers seriously and spam, left unchecked, would dilute the confidence our advertisers place in us. By being aggressive on combating spam, we want to make sure that doesn’t happen.”

To me this seems like a golden opportunity for Google. If they can figure out how to filter out the splogs, not only will it make AdSense advertisers happy, but it will also make their search results much better. I can’t understand why Google seems to need to be dragged into this fight instead of leading it.

The author of the story, Michael Pollitt, also has additional commentary on his blog.

Found via Spam Chongqing