nobarking’s posterous

I love a food fight when actual food is not involved. If the topic is Google Knol and quantity vs. quality is the issue, even better. This is my opening episode.

Thanks to the actions of one Knol creator -- in fact, the most prolific Knol creator -- fight ON. And I must thank my allies for jumping in without hesitation.

Scenario for the fight: an unverified fella named Rao who claims to be a University professor in India recently awarded himself "Top Author" of Google's Knol website. With thousands of writers and several hundred thousand articles or Knols to choose from, that is a high honor, indeed.

Note that I said "awarded himself." That's right. Google did not make this award. How and why did Rao do it?

First, at a Knol that purports to be a "foundation" for Knol authors but is really a private narcissistic conversation taking place in public view, Rao told his colleagues that he is now the global number one individual English Knol author, linking to an acceptance speech in the form of a Knol entitled Global Number One Individual English Knol Author. Rao's self title.

Quoting Rao, "Based on the page view statistics of 3.11.2009, the top ranking individual author position shifted to Narayana Rao K.V.S.S.

NOT!

Author rank at Google Knol is not a function of easily-gamed page views. Easy is Rao's game. This guy has "written" over 1,700 Knols + collections. That's more than three a day, seven days a week nonstop since Knol's launch in July 2008. And for all that, Mr. Rao's average Knol enjoys a mere 304 page views, including those prepared over a year ago. I'll be dealing more with "page views per knol" in another episode. For now, let it suffice that Rao is impressed by the fact that 304 multiplied by 1,700 produces a number just over one half million.  "How great am I that more than one-half million have read my knols."

By comparison. I have written an anemic 20 Knols + collections. But my average Knol enjoys 4,434 page views. My total page views is a mere 88,675. I am so far behind... or am I?

Also by comparison, writer Kevin Spaulding has penned 35 Knols, and his average Knol enjoys 13,715 page views. His total: 480,023. Just one of his Knols has earned 261,000 page views.

In perspective, Spaulding's average Knol pulls in more than 4,000% higher page views than Rao's average Knol.

For several months this year, Google rated me the #1 Top Author. The search giant figures this out (like everything else) with an algorithm. At present, my colleague Peter Baskerville in Australia is the Top Author.

Google is proud of its algorithm. It includes a direct link to it plus a photo gallery on Knol's home page. It is titled Top Author. That's Peter at the top. I'm down the page in position 7. Note that there is a bug in the algorithm. Positions 3, 8 and 9 are occupied by people who have written nothing. Position 4 is a massive plagiarizer. Google is working on the problems (they always encounter problems as people try to game the system).

Rao cannot be found in this rating, which produces a list of more than 400 authors. That must have made him angry enough to create his own recognition and even crow about it as if it is real. But, it is actually worse than the implied self-delusion. Rao is now encouraging others to do the same, to fill up Knol with thousands of articles. No mention of quality (why would he when there is no quality in his body of work -- it is mostly copy and paste stuff from old sources).

OK, episode complete for now. A healthy debate is raging about this (Rao has sent out a call for support) and I'll report as it develops.

Loading mentions Retweet

Today, Knol added a "highest rated" tab on the homepage that showcases newly created knols that have earned high marks from readers. You can select one month or all time. HOWEVER, one month is the default. It delivers writer Donald DePamphilis and only DePamphilis for 11 Knols that happen to have around 200 page views each but far more ratings than average for such low PVs

DePamphis is a good writer. He's a professor of economics. He seems to be using Knol to place case studies on-line. These 11 Knols look like the kind of stuff a professor might employ in a classroom. It's reasonable to suggest that students have been directed to these Knols as homework assignments. And that would explain the outlier stats on star ratings. Brown nosers!

I think it embarrasses Knol to dish up a new feature that is clearly broken from the get-go. Looking at Knol Help's current response history, it also seems futile to point this out to them. There are numerous serious unanswered complaints many days old that are ahead of this.

Is Knol sort of a stepchild application now? Probably not. Why would the engineers add a new feature if the service is on autopilot? However, Knol Help seems to have left the building. The engineers are in charge and no one is actually testing things before deployment, it seems.

Summarizing organically, imagine this scene. A new reader arrives at Knol's landing page, sees the tabs and selects "highest rated." Instantly, the screen fills with a list of Knols, all by one writer, all with the words Case Study in the title. Hmmm, only one writer? None of these titles interest me. Oh, well, Bye, Knol.

I'm embarrassed for Knol.

Loading mentions Retweet

Only treats in this batch of algorithmic winners for the month of October, 2009. Take a look, try a read. One of the winners is a high school student who felt compelled to share "How to Write an Effective Essay" with his peers. Another winner is my colleague Peter Baskerville for his enchanting history of Australia. Congratulations, winners.

http://knol.google.com/k/peter-baskerville/top-pick-best-knols-of-the-month/14j3i4hyjvi88/60#

Loading mentions Retweet

Just posted the following to members of the Google Knol group at LinkedIn. I am one of three managers of the growing group.

Knol is receiving new criticism in key blogs. Latest: "Google Knol Goes Quiet Next to Hubpages and Squidoo ( http://webtrends.about.com/b/2009/10/30/google-knol-goes-quiet-next-to-hubpages-and-squidoo.htm )

As Knol once again spins its wheels on stalled author page views and Google Help repeats old behavior, failing to respond (for five days now) to queries about this and other issues, members of this group have reason to worry. Or do we? How can we help Knol survive and flourish with Knols like Peter's "Australia," Garry's "Canina" and thousands of other worthwhile units of knowledge.

Does Google even want help? Twitter suggests YES, they value authors. But the lack of Help response suggests NO, once again Google is out of the building.

Daniel Nations, the writer of the above referenced article at About.com (I have a standing offer to become an About "guide" but convicts in prison are paid more), says "If we are to learn anything from Google Knol, it is that Google likes to experiment, but they don't always hit home runs. Google Knol joins such high profile experiments as Orkut, a social network popular in Brazil that never caught on in the U.S., and Google Answers, which couldn't quite capture the popularity of Yahoo Answers. So while new technologies such as Google Voice, Google Wave and Google Maps Navigation catch a lot of web buzz, there is certainly no guarantee that they'll end up being a big hit with users."

We know better in this group. Knol can be a big hit and that's why you joined this group. In my opinion, Knol needs a vaccine, an injection of Internet testosterone. Instead of letting Knol drift along passively, Google could take steps that would power purposeful navigation to the desired destination.

Tell me, tell the group what those steps might be in your opinion. I'll start:

1. Purge Knols over 6 months old that have failed to achieve more than 2 stars. This includes Knols that have failed to achieve any stars. And, announce the action in Google's blog.

Loading mentions Retweet

See the next story

Loading mentions Retweet

Glad to report that Peter Baskerville and I have completed a Knol that uses Google's algorithms to award Top Pick to the best Knols of the month.

Our research traveled backward to pre-launch of Knol in July, 2008. For each month, we figured out Gold, Silver, Bronze and Honorable Mention awards. The algorithms we used were designed by Google, not ourselves. These algorithms are secret, but employ page views, star ratings, comments and reviews to place Knols in rank order.

We were gratified to learn that our award Knol earned one of the awards.

The results are here: http://knol.google.com/k/murry-shohat/top-pick-best-knols-of-the-month/14j3i4hyjvi88/60#

As you can see, dozens of authors have been recognized for good work. As a final gesture, we split up the months and placed congratulatory messages on the winning Knols. Several of the winners have written back to thank us, and one winner has blogged about his success. We encourage more of this, as good work begets more good work. How nice it would be to have 10,000 people doing their best at Knol to earn one of these awards.

Given the rapid increase in page views for this Knol, Peter and I are thinking about adding Best Knols of the Quarter and Best Knols of the Year. Opinions?

This Knol also harvests the mashup between Google Knol and Google Docs to simplify and automate the presentation. Each month is a separate Doc created by copying a Master template that already contains a formatted color table. Because of the mashup, it's easy to copy-and-paste algorithm results into respective cells in the table.

Loading mentions Retweet

Just cranked up a new, team-written Knol to augment a major breakthrough at Google's Knol experiment. If you are concerned about Swine Flu, you may want to bookmark it because it could become "hot" as these things go. Just click the title:

H1Ni Influenza Real Time News Aggregation

Now, to the major breakthrough. After several months of planning, Google and the Public Library of Science (PLoS) announced a collaborative effort to instantly publish the fruits of fresh research about swine flu. The effort is named PLoS Currents: Influenza, and it lives as a "collection" at Google Knol. In the two days since it was announced, quite a bit of positive news has been generated. Here are the chief reasons:

• This "mashup" harnesses PLoS' commitment to open access for science information to Knol's instantaneous delivery model and implementation of a CCA (Creative Comments Attribution) license. In the context of the swine flu, this is ideal because it short circuits the usual slow and burdensome route for scientific publishing. My writing teammate, Dr. Krishan Maggon, is a biological scientist and he can't stop clapping about this.
• Since the PLoS folks want their collection to serve "for the rapid exchange of scientific results and ideas," we sensed a need to crossover from science to sociology. The rapid exchange idea should not stop at the University or research lab's doorway. In just two days, the PLoS collection has garnered thousands of page views. Our group would like to multiply those numbers by opening the University door. Our method is to aggregate related real-time news (and also provide some journalistic interpretation of PLoS content).

The PLoS collection employs a "vetting" technique based on expert collaborators. Articles in the collection have thus achieved sort of a peer review -- not full Journal-level review -- but sufficient testing to merit inclusion given the urgency of swine flu information distribution. This is really no different from the vetting offered by a magazine's carefully-chosen editorial board. But it sure is faster. Given my expressed concern in this blog about declining quality at Knol, we may be looking at a solution. Three words: expert-moderated collections. Thank you, Internet!

Yeah, I know. The gatekeeper theory returns. News is what the managing editor says is it. An old story. And a very good one. Given the decline in journalism at the print level, a new gatekeeper theory is completely appropriate. This one just might save Knol.

Loading mentions Retweet

I'm humored today. Is humor an appropriate emotion for the following situation? I'm humored after being attacked for performing research. And get this. The attacker is the creator of an anti-Wikipedia website for "non-notables." More on that in a minute.

The attack is directed at my efforts in the struggle over Will Johnson's Craigslist-like Knols, detailed in my two prior blog installments. The attacker does not like the fact that I've taken about an hour to investigate the flow of new Knols to determine if more Craigslist-like publishing has resulted from the publicity surrounding this event. To this, I say if not me, who; if not now, when?

Hospitals are for patients. Craigslist is for classified-style ads of all kinds. Knol is for units of knowledge by quality-minded writers. Arguably, a unit should be an informative article that help humans. Helping humans by selling used (or new) audio gear qualifies as a unit of knowledge about as much as my physical condition qualifies for the Olympics.

If Google does not police its own structure for creating user content, then it's entirely fair and reasonable for users to do it. Recognizing this, Google encourages users to do it on every Knol. I'm a user. Google's flagging, commenting, rating and reviewing tools are there for me and for you. In spite of the rantings of my attacker. In fact, his ability to attack is enabled by these very tools.

I'm not linking to this guy or his website because in my opinion he does not deserve your clicks. The website he set up -- a Wikipedia look-alike -- is for refugees from the Wikipedia world who cannot climb the ladder of notability to achieve a Wikipedia listing for themselves or their companies. In other words, wannabes for whom research puts the lie to notability. This is the Internet's lowest common denominator phenomena at work.

Surf around this guy's site and you'll see love for the non-noteworthy, persuasive appeals to greed and the announcement that the site owner is going to take some pennies from anything you contribute (but you are welcome to the rest... if you are able to get traction for it).  Why not write about non-notable people, companies and things by the ton? Write about Craigslist's new breed of direct advertisers, for example. Link to their ads. Lots of income potential for you!

NOT. Income for the site owner, lugnuts. Can lowest common denominator content achieve high search engine rank? I doubt it. And I must retract an earlier recommendation I made to one Knol'er to take a look at this site. I made it before taking a hard look myself.

Is Knol getting better or worse now that the cat is out of the bag about direct ads? Have you encountered any blatant ads yet? Here's one with a Craigslist-style title that leads you away from Knol (unlike the Johnson ad, which attempts to close you without leaving Knol until you are ready to pay. Isn't this just what all of us want:

Loading mentions Retweet

The breaking news of Knol's fall from QuantCast grace -- brought about by the revelation that Craigslist-style ads are beginning to appear as Knols, has filled the airwaves with mis- and dis-information.

TechCrunch's breaking story last night stimulated dozens of smarmy comments. Few of them contained actual wisdom anchored by analysis. Oddly, one of the latest comments, by the Craigslist ad perpetrator himself, was more insightful.

Will Johnson basically argues that we should follow the money. Google would much rather publish his kind of content (in this case audio speakers for sale) than host Knols about academic subjects. At least the Panasonics, Bose' and JBLs of the world might purchase Google ads to appear in the related content. Who would ever purchase an ad to appear in an article about "the spin of the electron or the meaning of Kant" asks Johnson. (Don't tempt me to answer; I have a bigger fish to fry).

Insight is oh so evanescent, Will. Let's hope Google is operating on a different frequency. Otherwise, thousands of authors will be betrayed. For those who take a little time to actually read about Knol, they'll learn that Google ramped the launch by recruiting armies of university talent to pen an inaugural set of articles (Knols). These are about some of the most pressing medical and dental conditions facing humanity, and a lot more. And this is exactly how I discovered Knol. I was doing research on Leukemia. Knol acquired at least three significant articles -- peer review worthy -- on blood cancers and bone marrow transplantation, and they showed up in my browser. I joined Knol immediately, on July 24, 2008.

These legacy authors include noted research professors, MD's, DMD's, DDS's and Ph.D's from top schools. Most of them have earned a ton of page views plus quality awards as top picks. For example, the Chairman of Tufts University's School of Dentistry, Dr. Noshir Mehta, is the top ranked English-language author at Knol. He's penned four popular articles on tooth pain and TMJ, hitting well over 100,000 page views. Together with a growing library of Knols by his university colleagues around the world, this legacy body of work greeted the public upon launch. It also demonstrated the full rich meaning of "a unit of knowledge." Google's launch slogan and elevator speech for Knol is absolutely clear. Today, these Knols hunt extremely well in search. Just Google "tooth pain" for a nibble.

By Johnson's argument, an ad for Bose bookshelf speakers is a better deal for Google than any of this academic stuff. Professors of physics and Kantian philosophy, stay away! Based on Google's behavior following the public launch, Will has evidence (some of it self-produced). After all -- and in spite of the care that went into the pre-launch library -- Google has allowed nearly anything and everything to become a Knol. Throwing caution to the wind, Google's content policy is so vague that Will Johnson believes with utter certainty that an ad to sell speakers is "a unit of knowledge." By that logic, Craigslist is the greatest repository of knowledge the world has ever seen, and is not likely to be dislodged anytime soon. I wonder what they know about Leukemia?

OK, I've stated the obvious. However, if the Quantcast data is right, Knol is failing. Craigslist-like ads just pour gasoline on the problem. And if Knol fails, thousands of legacy and street authors will be betrayed by Google. I am one of those street authors. Like the legacy guys, I've put my a-- into writing Knols based on deep research and workmanlike conduct. I'm going to be mighty fried if Google pulls that plug. There are several thousand Knol writers like me.

What will it take to save Knol? I've been getting the question all day. Two things leap to mind:

1. Set up an algorithm that automatically blocks from public view any Knol that fails to achieve a 3-star reader rating within 120 days. No traction, no publication. This just might begin to reverse the Quantcast data as readers learn that they no longer have to hold their nose due to the stench of a rotting mountain of garbage. There are thousands of high rank Knols in the system, written by earnest authors who grok the meaning of "a unit of knowledge."
2. Add explicit language to Content Policy to outlaw Knols that are ads. By saying yes to certain "commercial activity," Google is sending mixed and misunderstood signals to naiive writers who continue to believe in a free lunch. Knol's administrative team removes Knols that get flagged all the time. Will Johnson's Craigslist-like Knols (several ads) remain in place even though each has been flagged multiple times. Obviously, they've decided to turn this incident into "research." Or, they are waiting to see if Bose will buy a speaker ad for insertion with the Johnson Knols.
   

Loading mentions Retweet

My regular colleagues don't need to be reminded that I am a BIG fan of Google's Knol experiment. Today, as I quickly reviewed some of the newest Knols, my eyes did a double take on Will Johnson's Craigslist ad.

You heard me right. Click that link and take a look, or just peek at the picture here. Except for the fact that the content is a Knol or article dished up by Google as a "unit of knowledge," this is exactly like a Craigslist ad. Right down to the photo and offer.

Knols as classified ads? Not on my watch. I immediately "flagged" the Knol, complaining to Google about clear content policy violations. And I left a comment for Will. Then, an investigation commenced. Here are my discoveries:

• Will has nine Knols like this, and more links selling other things off-site (e.g., at Amazon.com)
• One of Will's Knols is titled My Knol Marketplace. Sort of a super ad of other ads. Except, not super at all.
  
• These Knols violate Google's content policy on several levels. Since I'm sort of the chief investigator for my team Knol on Content Violations, my detective hat went on and off came the gloves. Will's work is now the active Case in Point.
  

Using Google's own search tools, I hunted for others like Will. People who are using Knol as their personal storefront, market or bodega. I have failed so far to find anyone as bold as Will. In fact I have failed to find any comparable Knols. Time has been brief for this search, so I'll try harder. Nevertheless, it speaks for itself that Knol is a website for "a unit of knowledge."  Is Will representative of a very small group that doesn't get it? Ads are NOT units of knowledge. They are ads. What don't you get?

There's plenty of room for confusion about content in Google's content policy. That why my team writes about Content Violations. Google does not want to be in the editorial business. So their broad-stroke content policy is really a set of guidelines for the creation of public policy by consensus. That's what my team believes. Public policy requires proactivity. That's a role we've accepted, and based on quality badges, reader comments and reviews, our work is appreciated. Occasionally we'll hear directly from Google, and it's almost always in support.

So do me a small favor. Go to Knol, click Will's My Knol Marketplace (link above), and FLAG every one of his Knols. When you do, you are offered some choices. I recommend that you select Other, and write the following comment in the box (in your own words): Craigslist on Knol? What are you guys thinking? Dump this author.

Loading mentions Retweet