Home

European conference on machine learning and principles and practice of knowledge discovery in databases (ECML/PKDD 2010)

Mon, 15/02/2010 - 8:00am | by daniel

[apologies if you receive multiple copies]



The focus of the ECML/PKDD 2010 Discovery Challenge is Web Content

Quality. High quality is not simply the opposite of Web Spam. The recent

Web Spam Challenges have explored the aspects of filtering as a binary

decision. In this year's Discovery Challenge we target at more and

different aspects. We want to develop site-level classification for the

genre of the web sites (editorial, news, commercial, educational, "deep

Web", or Web spam and more) as well as their readability,

authoritativeness, trustworthiness and neutrality.



The data set will consist of sample Web hosts from Europe in three

languages (English, French and German). The training and testing samples

will be biased towards the interesting aspects and cleansed manually

from mixed sites, Web hosting, and adult content. Features similar to

those used to filter Web spam based on content and linkage information

will be provided on the host level, along with natural language

[More...]

See original: Eventseer.net - the academic event tracker European conference on machine learning and principles and practice of knowledge discovery in databases (ECML/PKDD 2010)

Why Fish Hold the Key to Increasing Wind Farm Power

Mon, 15/02/2010 - 6:10am | by daniel

The way schooling fish swim reveals how to squeeze more power from the wind over a given area of land.

What do wind farms and fish schools have in common? Not enough, according to an analysis by Bob Whittlesey and buddies at the California Institute of Technology in Pasadena. And here's why.
Most wind farms use horizontal axis turbines because they generate significantly more power from the air flowing through them than vertical axis turbines. That's why you seen plenty of wind farms using horizontal axis turbines and very few using vertical axes.
But there's a problem with horizontal axis turbines. The maximum power a rotor can extract is proportional to the cube

of the windspeed. But sinc, they slow the air passing through them, neighbouring turbines need to be around ten turbine diameters apart. And that places a significant limit on the power that can be generated from a given area of land.

The anecdotal evidence is that vertical axis turbines do better--by some accounts closely packing these turbines can actually increase their power output. Today Whittlesey and co explain why.
The secret appears to be a well known aerodynamic effect called a Karman vortex street, a term you may not know but will surely be familiar with from the example below, wich shows the cloud patterns downwind of the Juan Fernandez Islands off the coast of Chile.
These repeating patterns of vortices in an air stream around a body are relatively common. In fact, fluid dynamicists have well developed models for simulating the way their interaction helps keep schools of fish synchronised and reduces the total propulsive power needed per fish. A similar effect reduces the fuel consumption of vehicles travelling in a platoon.
So for the Caltech kids, it was but a small matter to redesign such a model to simulate the vortices in a 16 x 16 farm of vertical axis wind turbines and to study the effect of close packing the turbines.
It turns out that the additional acceleration that the vortices add to the wind can increase the power of vertical axis turbines. And although they still do not have the power efficiency of horizontal axis turbines, this means they can be packed much more closely together.
That's important because it increases the power that can be genetated from a given area of land and not by a small amount. "These configurations significantly reduced the land use for vertical axis wind turbine wind farms, resulting in array power density increases of over one order of magnitude compared to operational horizontal axis wind turbine wind farms," say the Caltech boys.
Yep, an order of magnitude increase in power density.
Of course, the model is only two dimensional and does not account for complex flow behaviour such as vortex shedding. But is it worth looking into in more detail? You betcha. Caltech has already filed patents.
Not bad for an idea inspired by fish.
Ref: arxiv.org/abs/1002.2250: Fish schooling as a basis for vertical axis wind turbine farm design






See original: Technology Review Feed - arXiv blog Why Fish Hold the Key to Increasing Wind Farm Power

Very young people blogging about science - let's welcome them [A Blog Around The Clock]

Mon, 15/02/2010 - 2:18am | by daniel

A few days ago, I asked what it takes for a young person to start and, more importantly, continue for a longer term, to write a science blog. The comment thread on that post is quite enlightening, I have to say - check it out.

What is more important - that post started a chain-reaction on Twitter and blogs. Arikia Millikan, herself a young blogger, wrote a post in response which also attracted a lot of interesting comments. Go and comment.

Mason Posner wrote not one, but two posts in response: Science blogging in the classroom, an update and Young science bloggers need community. Go and comment.

Some of his students also congregated on his Facebook wall and, energized by all the spotlight they were getting, decided to restart their old class blog: Science Haggis. Go and comment.

Amy Breslin, former student of Posner, is the only one of his last year's students to have continuously blogged ever since, on Plague-erism. Go and comment.

Then, someone on Twitter brought this link into the discussion - a blog post by a science blogger on The Life of Pi explaining one's own insecurities about blogging and why it is hard. Once that link got passed around on Twitter by a bunch of people, the blog post received a lot of encouraging and wise comments as well. Go and comment.

Christie Wilcox, who is a better known youg science blogger, also voiced some similar uncertainties after coming back home from ScienceOnline2010. Go and comment.

What many of these blog posts and comments point out is that it is really hard to keep blogging if the audience is invisible. It is an absolutely astonishing coincidence that Anil Dash wrote a fantastic blog post on exactly the same topic just yesterday.

The current technology online makes it easy for you to see who you follow and read. It makes it easy on some platforms for others to see who you follow and read. But it is almost impossible to see who is reading you! Where's the audience? Am I just blowin' in the wind?

In a way, traditional blogging, in the absence of much feedback, is a one-to-many communication, which is not the best way to do it.

Sure, you can use various software to see how many people subscribe to your blog feed - but not who they are or if they are reading you at all. You can find out many bloggers who put your blog on their blogroll - but you still don't know if they are actually reading you.

There are two ways people can tell you if they are reading you. One is to link to an individual post of your (not just the homepage). A simple link with no commentary on their blog or Facebook or Twitter or FriendFeed etc., is a simple statement "this may be interesting to you" targeted at their audience - it does not mean endorsement, but it is nice nonetheless. A link that adds commentary to it - agreement or disareement or addition of further information or providing an additional angle - is even better. You can find the links to you in your tracking software (Sitemeter referrers list, Google Analytics, etc.) or by putting your blog URL in Google Blogsearch or Technorati.

The other, much better way to let you know they read your post is to post a comment on it. Once they do - and posting the very first comment is the hardest - reply! Don't make commenting on your blog difficult or exclusionary. Keep it open. You will get a substantive, pleasant discussion in the comments if a) you set the tone in your own post, b) carefully monitor the comments, c) moderate as needed, and d) respond frequently. Do not make the mistake that newspapers made of letting the loudest, most obnoxious commenters take over and scare away everyone else. At the same time, do not quickly delete every comment the tone of which you don't like - this also has a censoring effect and will not make you many friends. Make your own criteria, draw your own line.

So, the best way to encourage a blogger - any blogger, but especially a new or young one - is to post comments. Good, quality comments. You may be used to the Usenet tone, but n00bs take some time to get used to it. Be gentle toward the young 'uns. Go and comment.

For the new bloggers - of course there is some advice (including that already mentioned in the many comments on the blogs I linked above).

If you write a post about a peer-reviewed paper - have it aggregated on ResearchBlogging.org: this will bring yo not just traffic, but also respect. Not everyone can have their stuff up there - you need to apply and get approved first.

Send your best posts to blog carnivals on a regular basis. You'll get traffic, new readers and will be joining a community of bloggers interested in the same topic.

Shameless self-promotion is not a bad word any more. In the world of the Web, nobody will know your blog exists unless you say "Here I am - look at me!" sometimes (yes, keep it tasteful, but it is OK).

Comment on other blogs and use your blog URL as a link that people will follow when they click on your name. The blog owner is almost certainly going to click there.

Link to your best recent posts on other online platforms: Facebook, FriendFeed, Twitter, etc. E-mail the link to your Mom every now and then. Marketing yourself has become an essential aspect of communication in the 21st century - nobody will do it for you any more.

Here are some other new/young bloggers of note:

Naked Little Ape is a blog by Hannah Lucy King. The discussion of this topic on Twitter persuaded her to make her blog public and to promote it there. And the blog is fascinating! Go and comment.

The Difference between Ignorance and Apathy is one of the current student blogs in Posner's class. Go and comment.

SexyScience is one of the current student blogs in Posner's class. Go and comment.

Thirsty Pandas is one of the current student blogs in Posner's class. Go and comment.

Successors of Solomon is one of the current student blogs in Posner's class. Go and comment.

Trisha Saha is the only one from the Duke Summer class who continued blogging after the course was over. And even she has not posted in a while. Bloggers on Nature Network have no access to tracking and traffic statistics, so the only way she can possibly know if someone is reading is if someone posts comments. Perhaps she will blog again if she starts getting comments on her older stuff. Go and comment.

Anne-Marie Hodge, though so young, is already a veteran science blogger. Since moving from undergraduate to graduate school she is busy and her blogging has become more infrequent. Though, when she posts it's awesome. She is also on Nature Network so the only way you can make invisible audience become visible to her is if you post comments. Go and comment.

Miss Baker's high school biology students are posting on Expert Biology. Check out Jack's, Ammar's and Alex's posts about ScienceOnline2010. Check their other posts. Go and comment.

Lauren Rugani is a young science blogger/journalist. Go and comment.

Christine Ottery is a young science blogger/journalist. Go and comment.

Elissa Hoffman's students are also blogging. Go and comment.

Dale Basler's students are blogging. Go and comment.

Naon Tiotami is a very young blogger. Go and comment.

Sam Dupuis is a very young blogger. So is Djordje Jeremic (see this). Go and comment.

Mimi is a wonderful young blogger. Go and comment.

Students are blogging on the Project Exploration Blog. Remember Project Exploration? This is where it all started. Go and comment.

Let's make sure new and young science bloggers feel welcome in our community. Let's help them make their audience visible. Go and comment.

Read the comments on this post...

Also check out the featured ScienceBlog of the week: Laelaps

See original: ScienceBlogs Select Very young people blogging about science - let's welcome them [A Blog Around The Clock]

Leakegate: the case for fraud [Deltoid]

Mon, 15/02/2010 - 1:37am | by daniel

They have been some explosive new revelations in the Leakegate scandal. Remember how Leake deliberately concealed the fact that Dan Nepstad, the author of the 1999 Nature paper cited as evidence for the IPCC statement about the vulnerability of the Amazon had replied to Leake's query and informed him the claim was correct? Leake didn't report what Nepstad told him. Instead he claimed that the IPCC statement was "bogus", even though he knew it wasn't.

Deltoid can now reveal that Leake's reporting was far more dishonest than originally believed. This is how Leake quoted Simon Lewis:

Simon Lewis, a Royal Society research fellow at Leeds University who specialises in tropical forest ecology, described the section of Rowell and Moore's report predicting the potential destruction of large swathes of rainforest as "a mess".

"The Nature paper is about the interactions of logging damage, fire and periodic droughts, all extremely important in understanding the vulnerability of Amazon forest to drought, but is not related to the vulnerability of these forests to reductions in rainfall," he said.

But Lewis, like Nepstad, told Leake that the IPCC was correct. When asked whether Leake had represented his views accurately he replied:

Absolutely not. Please see the BBC piece online, by Roger Harrabin. I sent them the same email with my comments. The scientific statement in the IPCC WG2 report is essentially correct, but has a referencing error. IPCC WG1 get it right. An outrageous piece of journalism.

Read the rest of this post... |

Read the comments on this post...

Also check out the featured ScienceBlog of the week: Laelaps

See original: ScienceBlogs Select Leakegate: the case for fraud [Deltoid]

Kandel on camera [Neuron Culture]

Mon, 15/02/2010 - 1:03am | by daniel

Kandel

I profiled neuroscientist Eric Kandel for Scientific American Mind a while back; a huge pleasure. Two things stand out.

 First, Kandel's work makes a wonderful foundation for an understanding of neuroscience, as his mid-20th-century insights into the dynamics of memory underlie much of the discipline.

Second, Kandel  is a gas -- gracious, funny, and stunningly brilliant. When I interviewed him for about 90 minutes in his office at Columbia, he was 73. As he described to me the history of his work, and of modern neuroscience, he seemed to have complete and effortless recall about everything. If he talked about a finding, he would remember everyone who worked on the paper, when the paper was published and where -- and where it was in his vast file cabinets. The master of memory has MEMORY.

And his talk flowed effortlessly. At one point we were interrupted by an assistant who came in to review an illustration that was being prepared for Kandel's memoir, "In Search of Memory," which he was then finishing. Kandel answered the questions clearly, telling the assistant where to find the information needed and so on. Took about 5 minutes. The assistant left. As he closed the door, Kandel turned back to me, said, "Excuse me. As I was saying ..." and then, to my astonishment (I was glancing at my notes to see where we'd left off) resumed talking by returning to the beginning of the sentence he'd been in the midst of when the assistant entered the room. (Later, when I was telling my wife about this, I said, "The man is smart." She said, "That would explain the Nobel Prize."

I highly recommend http://www.amazon.com/gp/product/0393329372?ie=UTF8&tag=daviddobbs-20&li..., which is a splendid picture of a rich life devoted to science -- and a great review of much of 20th-century neuroscience. If the film catches even a fraction of his accomplishment and charm, it should be a real treat. From the Times:

It's not often that you are invited to spend an hour or two in the presence of a Nobel Prize winner, and "In Search of Memory: The Neuroscientist Eric Kandel," Petra Seeger's new documentary, offers an especially gratifying opportunity.

via movies.nytimes.com

image: Icarus Films

Posted via web from David Dobbs's Somatic Marker

Read the comments on this post...

Also check out the featured ScienceBlog of the week: Laelaps

See original: ScienceBlogs Select Kandel on camera [Neuron Culture]

Why is 'scientists are bad communicators' trope wrong [A Blog Around The Clock]

Sun, 14/02/2010 - 10:05pm | by daniel

For a very long time, I have argued that many scientists are excellent communicators.

I have seen a number of scientists talk over the years and the experience has been mostly very positive. Even if I limit myself only to what I saw over the last couple of months, every single scientist lecture was riveting.

So, where does the "scientists are bad communicators" trope come from?

I think it comes from the people looking at the results - a country whose government (and population) does anti-scientific stuff. They look at various factors that may lead to that state and decide that the audience, while uninformed, is interested in science; that science education is too difficult to fix; that movies portray scientists in a bad light (which may be wrong); that the media does not cover science enough, etc. How do they deduce from this that if only scientists could talk better we can make progress, I don't know.

I have written at length (I know it's long, but it's worth reading) a critique of this conclusion. There are not enough scientists to, even if they were all brilliant speakers and spoke every day, make any difference. The problem is with the "push" versus "pull" models of communication. Many scientists communicate well, but are only allowed by the mainstream media to use the "pull" model which attracts only those who are already interested in science. The examples of "pull" media for science are popular science magazines, news sections of scientific journals, science sections of newspapers, science blogs, science-related radio shows, science-related shows on cable TV, i.e., all those places where people have a choice to seek this information or bypass it.

It is the mainstream media that controls all the "push" venues - the most popular print, radio and TV venues that are seen by everyone and where science could, potentially, be mixed in with the news coverage of other areas of life, thus delivering science stories to people who otherwise would never seek them. And it is there that the scientists have no access, certainly no access on their own terms, and thus it is there where the science communication is blocked. Scientists communicate all the time, and do it well, but only to the already receptive audience which actively seeks them - in special sections, or self-made media, carefully quarantined away from the mainstream news. The corporate media actively prevents the scientists from access to the non-receptive yet potentially interested audience. Thus, it is no surprise that some of the purveyors of the "scientists are bad communicators" trope are themselves journalists, parts of the corporate media culture and thus oblivious to the ways their own professions hinders the communication of science (and thus building trust in scientists) to the masses.

I am not the only one to think so.

But there is another reason why some people accept and push the "scientists are bad communicators" trope. Their understanding of communication - what it is and how it works - is out-dated. It is pre-Web, and they do not grok how the Web changed everything. All the academic literature on communication published earlier than late 1990s is now useless: not just outdated, but wrong.

@DrPetra said it succintly on Twitter the other day:

the 'scientists are bad communicators' still implies some one-off talk/top down approach. Public engagement = a dialogue

And this is the key. The "scientists are bad communicators" trope requires thinking in a one-to-many mode of communication. It is stuck in the mid-20th century way of thinking about science communication: the scientists give lectures, science cafes, write popular articles, perhaps a Sagan-wannabe shows up on TV. All of that is one-to-many. And all of that deals with communication in terms of "I am the expert, I talk, you listen". But, a couple of decades into the Web era, audience does not accept this mode of communication any more. This kind of communication does not increase but actually decreases the trust in the person who is doing the talking - "who is this haughty guy and who does he think he is to talk down to us and not listen to us or even let us respond?"

If you were at ScienceOnline2010 or watched it from afar, especially the media/journalism 'track' of conversations, you would have noticed that pretty much everyone there came to the same conclusion - the one-to-many model of communication is out-dated. It is a part of one's toolkit, but on its own it can potentially do more harm than good if one's goal is the popular trust in science and scientists.

The way to gain popular trust in science is not so much to communicate one's expertise to passive lay audience, as it is to engage. The other day I tweeted that I am at my best as a science communicator when I am answering someone's question on Aardvark. Why? Because it is social. It is a two-way street. Even more so than blogs or Twitter, because of technical inefficiencies in these platforms in 'seeing one's audience'.

So, while the ability to give a riveting talk is still a great talent to have (or at least something that can be practiced and made perfect), it is not just not enough - it ignores what is really important in gaining the respect and trust of the lay audience: and that is to find the un-interested lay audience and make them interested. The "push", not the "pull" (see clip).

How do you find and get attention of un-interested audience? You go where they are and engage, not lecture them. If you cannot get access to the mainstream media's hot spots, you go around them, to where the people are: online. On Facebook, FriendFeed, Twitter, LiveJournal, blogs, Google Buzz, aardvark, etc. Engage, don't preach. The same goes in the classrooms - don't give guest-lectures: engage the students in discussion, experiments, even Citizen Science.

The best public speakers, those who get invited to do one-to-many lectures, often diverge from the traditional model and insist on being interrupted with questions during the talk, and leave plenty of time for many questions afterward. This is also why an unconference is much more useful (and pleasant) and more effective than a traditional conference. Now that the people formerly known as audience can talk back, they expect to be given the opportunity to talk back and putting any barriers to this pisses them off - thus you fail as a communicator.

So, not understanding the modern principles of communication in the Web era and relying on outdated academic literature on communication pre-Web is not just outdated, but wrong. Teaching others about this kind of communication as if it was the latest thinking in the field is not just "oh well, outdated but won't hurt" - it actually hurts our cause! It teaches scientists, who are already good communicators, how to become worse at it. Instead of teaching them how to break out of the kabuki of science communication it teaches them how to get even more entrenched in it and to even more fiercely defend the kabuki and the academic formal hierarchy that the kabuki represents. This sets us all back.

Read the comments on this post...

Also check out the featured ScienceBlog of the week: Laelaps

See original: ScienceBlogs Select Why is 'scientists are bad communicators' trope wrong [A Blog Around The Clock]

IPCC errors: facts and spin

Sun, 14/02/2010 - 9:39pm | by daniel

IPCC AR4 errors, glaciergate, amazongate

See original: RealClimate IPCC errors: facts and spin

Every child is a casualty... [Greg Laden's Blog]

Sun, 14/02/2010 - 8:58pm | by daniel

... when it comes to Creationist Home Schooling.

Read the rest of this post... |

Read the comments on this post...

Also check out the featured ScienceBlog of the week: Laelaps

See original: ScienceBlogs Select Every child is a casualty... [Greg Laden's Blog]

Amy Bishop UAH case: What role should personality or collegiality play in tenure decisions? [Terra Sigillata]

Sun, 14/02/2010 - 7:02pm | by daniel

Valued commenter wc just left us a link to one of the most insightful articles to date on Dr. Amy Bishop, the University of Alabama in Huntsville biology professor charged in the shooting deaths of three colleagues where two other professors and an administrative assistant were injured.

In today's Decatur Daily, staff writer Eric Fleischauer has an extended interview with UAH psychology professor Eric Seemann. You really should read the whole thing because it provides an inside view of Bishop's personality and relationships. But here is a critical passage:

Despite her excellent research ability, Seemann was not surprised she struggled to obtain tenure.

"Amy was kind of hard to get along with," he said. "I've talked to people who said, 'Wow, she can be really arrogant,' or be really headstrong. I knew that to be true. But at the same time she was brilliant. She was really one of UAH's rising research stars. People I know in biological sciences would say, 'She's a great researcher, but she's lousy to work with.' "

She was brilliant and she knew it.

"At one meeting I was with Amy, she was complaining to a group of us. She said she was denied tenure not because she was a lousy researcher -- she's not, quite the opposite -- and not because she didn't have good classes, she believed she did -- I think some might say otherwise -- but because she was accused of being arrogant, aloof and superior. And she said, 'I am.'

I recently had the opportunity to lead an effort to draft from scratch a reappointment, promotion, and tenure document for a newly-established department. With a committee of deans and department chairs, the final document pretty much included your typical quantitative requirements for teaching, research, and service. But one dean strongly suggested to me that we include a section on collegiality, defined loosely as the ability to interact constructively with individuals for the greater good of the department and the university. While wording to that effect was included, it was not explicitly defined as an evaluative criterion.

In academia, we often tolerate a great deal of destructive and defiant behavior that disrupts the organization in the name of "genius," perceived external stature and, perhaps most importantly, grant dollars (which generate indirect cost dollars for the institution). When describing some situations I've encountered to my colleagues in other non-academic businesses, their conclusion was that some of these people would often be let go if such behavior occurred in their workplaces.

For this consideration, let us step away for a moment from the horrible tragedy in Huntsville. Let us assume that an assistant professor there adequately met all of the explicit quantitative criteria for promotion and tenure in terms of teaching, research, and service. I would expect, however, that if the candidate under consideration was not an otherwise constructive member of the organization, comments in this regard would have been included in the chair's recommendation to the college dean's promotion and tenure committee based on the deliberations of the departmental promotion and tenure committee.

The questions for you, dear academic reader are:

1. Do you think that lack of collegiality is grounds for denial of tenure for a candidate that otherwise meets the basic quantitative criteria outlined in university guidelines?

2. Do you feel that collegiality - or whatever you want to call it: teamwork, cooperation - should be an important factor in making academic tenure decisions?

Read the comments on this post...

Also check out the featured ScienceBlog of the week: Laelaps

See original: ScienceBlogs Select Amy Bishop UAH case: What role should personality or collegiality play in tenure decisions? [Terra Sigillata]