gpeerreview – Google Code

http://community.brighton.ac.uk/jd29/weblog/41119.html

Full story at: http://jondron.net/cofind/frshowresource.php?tid=5325&resid=1392

This is mighty cool. Mighty mighty cool…

“What is GPeerReview?

GPeerReview is a command-line tool that makes it simple to write a review of someone’s work and digitally sign them together.

How does it work?

1. First, you read someone’s paper.

2. Next, write a review. (The review is just a simple text file that contains a few scores and your opinions about the paper.)

3. Use GPeerReview to sign the review. (It will add a hash of the paper to your review, then it will use GPG to digitally sign the review.)

4. Send the signed review to the author. If the author likes the review, he/she will include it with his/her list of published works.

5. Prospective employers or other persons can easily verify that the reviews are valid.

Why?

* Peer reviews give credibility to an author’s work.

* Journals and conferences can use this tool to indicate acceptance of a paper.

* Researchers can also give credibility to each other by reviewing each others’ works.

* This enables researchers to publish first, and review later.

* It meshes seamlessly with existing publication venues. Even the credibility of works that have already been published can be enhanced by obtaining additional peer reviews.

* A decentralized social-network of reviewers and papers is naturally formed by this process. The structure of this network reflects that of the research community. “
Created:Thu, 05 Feb 2009 08:56:45 GMT

A GCSE module in 90 minutes, including 30 minutes of basketball

http://community.brighton.ac.uk/jd29/weblog/39700.html

http://www.guardian.co.uk/education/2009/jan/30/gcses-schools

A school in the UK has been testing spaced learning (a technique based on findings about memory formation from neuroscientific research) to condense a four-month GCSE science module into 90 minutes – 20 minutes of intensive narrated Powerpoint, 10 minutes of basketball, repeated three times, with great success. In fact, over a quarter of students did better in the tests using this method than they did after subsequently taking the traditional four-month module. The suggestion is that an entire GCSE could be passed by most students with just a few days of study and, strikingly, that further study might actually be harmful in a significant number of cases.

Ignoring things like the Hawthorne effect and assuming these results are meaningful, there are two main conclusions to be drawn here. The first is positive: that spaced learning works pretty well and that we can learn a lot from neuroscience. The second is appalling: that GCSEs (qualifications usually taken by English students at the age of 16) are almost totally useless as a means of gauging knowledge and understanding. Of course, we suspected that already.

Tests and exams are so embedded in our educational systems that we sometimes think they tell us something useful about the effectiveness of teaching and learning strategies. Alas, they tell us little. What they do tell us is that, somehow, a particular instance of a particular intervention may have helped some people to pass the test. If we get enough similar interventions in enough contexts to help identify a pattern, then we can start to say with some assurance whether a particular kind of intervention might help some people pass some kinds of test, and we might even be able to generalise a little about shared characteristics of such people, which might in turn help us to tailor our teaching for different learners to pass tests more effectively. Whether the test tells us anything useful or not, however, remains a significant question.

In this particular context, there is some evidence that spaced learning may be an effective approach to passing some GCSEs. But even here there are some nasty issues: the fact that many of the same students actually did worse after following this process and then studying for a further four months suggests that:

  1. whatever they learnt was not persistent and/or
  2. that what they learnt later reduced their ability to pass the exam.

If the former is true, spaced learning may have its uses but they are pretty limited. Given the advantages conferred by having already had a successful go at the tests, if the latter is true, then either it is a sign of some truly appalling teaching or, more likely, it suggests that the students carried on learning and may have subsequently known too much to pass. This sounds bizarre, but I have some anecdotal evidence for this: I can remember looking through model exam questions and answers for a GCSE-equivalent computing course with my son a few years ago and being horrified that he was being penalised for knowing too much on many of the questions, which frequently ignored complexities and ambiguities in favour of repeating what the book (sometimes absolutely wrongly) stated. For instance, one that stands out in my poor memory is that markers were explicitly told to penalise students for stating (correctly) that TCP and IP are protocols, while rewarding the incorrect answer of TCP/IP (which is actually a suite of protocols). A student with curiosity and an interest in the subject who had explored even a little further than the book would therefore have received lower marks than those who had memorised just what was needed. It is not surprising, therefore, that a relatively surface approach would be more successful in such instances. Knowing a little of the right kind of facts to answer test questions would, at least sometimes, be more useful than actually understanding the subject.

So, in the context of exam-passing at least, spaced learning is either useless in the long term, or part of the reason for its success is that it emphasises surface-level memory skills at the expense of depth of real learning. Interesting, but not revolutionary.

There are some occasions in life when this kind of learning can be useful (I'd like to try spaced learning as a means of learning to play a song, for instance) but not enough to warrant its wholesale adoption. More significantly, I think that it's yet another a damning indictment of tests/exams as the primary driver and means of evaluating the success of our educational system. There are huge opportunities to rethink what we are assessing and how we do it, and we must work on these urgently. Assessment is such a driver in our systems that, if we do it wrong, we run a big risk of setting inauthentic goals and encouraging weak learning strategies that must be unlearnt as we enter real life.

 

 

Jinni

http://community.brighton.ac.uk/jd29/weblog/39632.html

Full story at: http://jondron.net/cofind/frshowresource.php?tid=5325&resid=1391

Jinni claims to help you find movies, TV shows matching your taste watch online. I spent a few minutes training it and it’s not that great yet, but I do have eclectic tastes which might mess with its algorithms a bit and the system is obviously still growing.

What is interesting about it is its combination of expert opinions/classifications, machine intelligence (they talk of a movie genome that uses a rich ontology, akin to the music genome that led to Pandora) and collaborative filtering. It is clearly trying to marry the top down and bottom up in an interesting way. The model they seem to be using allows for the bottom-up to become more prominent as time goes by. I suspect that, as the user-base grows and the cold-start problem lessens, that this might turn out to be quite useful.

In its combination of sophisticated (and apparently recursive) algorithms and human input it is a fine example of a collective application. The use of two distinct strata of human input (the experts and the rest of us) gives an extra twist and a potentially richer dynamic than the usual fare.

Its use of an ontology offers benefits of parcellation as well as a richer set of ratings than the usual ‘this is good’ approach. In addition to the usual movie metadata, the main divisions are ‘experience’ and ‘story’, with each aspect subdivided into many other subtypes. The ‘experience’ aspect is particularly interesting, parallel in some respects to my own CoFIND system’s use of qualities, albeit in a more structured and less user-led form. The structure serves a purpose, though, allowing them to automate tagging once the system has been trained. If it works, this might help to overcome the problem of spiralling complexity and everlasting cold starts that have proved to be a stumbling block for CoFIND.

I look forward to seeing how this develops.
Created:Mon, 26 Jan 2009 18:10:57 GMT

49 Amazing Social Media, Web 2.0 And Internet Stats

http://community.brighton.ac.uk/jd29/weblog/39582.html

Full story at: http://jondron.net/cofind/frshowresource.php?tid=5325&resid=1390

The figures speak for themselves and are pretty much what I would have expected on the whole, but one or two caused me to do a double-take. This one surprised me: YouTube’s bandwidth costs per day are about $1,000,000. That’s $365m per year on handling 13 hours of uploaded video every minute and well over 100m videos viewed every day. I guess it doesn’t seem that expensive when you think of it that way.

Created:Sat, 24 Jan 2009 19:46:50 GMT

OpenSocial, OpenID, and OAuth: Oh, My!

http://community.brighton.ac.uk/jd29/weblog/39386.html

Full story at: http://jondron.net/cofind/frshowresource.php?tid=5325&resid=1389

A terrific talk by Joseph Parr of Plaxo. This video explains the technologies behind social sites very clearly. It’s an hour long but, if you’re interested in developing social applications and you’re not sure where to begin (or even worse, you *are* sure but haven’t heard of these standards) then it’s a great introduction.
Created:Wed, 14 Jan 2009 06:22:08 GMT

TagCrowd – make your own tag cloud from any text

http://community.brighton.ac.uk/jd29/weblog/39309.html

Full story at: http://jondron.net/cofind/frshowresource.php?tid=5325&resid=1388

A simpler, primitive but very usable and less over-technologised system than Wordle that takes some plain text (or HTML from ANY web page) and turns it into a tag cloud. I saw Wordle when it was relatively young (months, not years ago!) and it was slightly more like this, though even then had some novel output options and was less developer-focused than TagCrowd. TagCrowd generates very clear legible, standards compliant, HTML/CSS but little else. They profess a desire to build an API, but it has none yet. Even so, sometimes simple is beautiful. A nice little system.
Created:Sun, 11 Jan 2009 10:32:53 GMT

James Paul Gee on games, social media and education

http://community.brighton.ac.uk/jd29/weblog/39250.html

 http://www.edutopia.org/james-gee-games-learning-video

A marvellous video from Edutopia featuring James Paul Gee in which he presents some very persuasive arguments for games and social media in education. More importantly, he challenges how school education is done in the US (although there are local differences this is much the same as it is done most of the world when you get down to basics, and pretty much the same as much of university education, especially in the sciences) and offers some ways out. Not much is new in what he has to say, but he says it very well. Enjoy!

Happiness spreads better than sadness

http://community.brighton.ac.uk/jd29/weblog/39192.html

Full story at: http://jondron.net/cofind/frshowresource.php?tid=5325&resid=1387

A great study by James Fowler (the same James Fowler who discovered that obesity is infectious through social networks) and Nicholas Christakis. It seems that happiness ripples through a population. Thankfully, it ripples slightly more effectively than sadness.

Their conclusion:

“People’s happiness depends on the happiness of others with whom they are connected. This provides further justification for seeing happiness, like health, as a collective phenomenon.”

The research is an offshoot of the remarkable Framingham Heart Study, instigated in 1948 and carried on through generations of volunteers. The experimental methods seem to have effectively dismissed the possibility that the effects are a result of random clustering, homophily or confounding factors like joint experience of an economic downturn or neighbourhood upheaval.

The results are fascinating. We are 15% more likely to be happy if someone with whom we have a close connection is happy. The effect is greater than the unhappiness caused by unhappy close people. As a result, the better connected our friends and family are, the more likely it is that we will be happy:

“Happy people tend to be located in the centre of their local social networks and in large clusters of other happy people. The happiness of an individual is associated with the happiness of people up to three degrees removed in the social network.”

Becoming happy is a good thing for all concerned – an unhappy close friend becoming happy increases the chance that we will become happy by 25%. If the friendship is reciprocal, it increases the effect by 63%. Interestingly, friends and next-door neighbours have more effect than spouses, which may partly be explained by the fact that happiness spreads faster through same-sex relationships (they don’t discuss gay relationships though!).

Physical proximity is very important – the effect decays noticeably, even between next door neighbours and those a few doors away. As the study looks at data from 1971-2003 it is hard to draw any conclusions about the effects of computer-mediated relationships and the authors are careful to point out that they can only speculate on the mechanisms for transmission. It could be anything from the effects of happiness on behaviour (generosity, helpfulness etc) to the direct effects of smiling, to the influence of pheromones.

The authors observe that educated people are generally happier than those who are not (and, incidentally, that women are generally sadder than men, but that’s another issue). I’ve wondered in the past about how we could adopt an infection model of education. This gives another driver that could make it work. Happy people are generally better educated and happy people get better connected. So education can have a disproportionate effect on happiness that goes beyond the benefits to the educated. This is good news all round and further proof (if it were needed) that the absurd notion of treating students as customers should be relegated to history. The customer of education, it bears repeating, is society, not the student.

Created:Wed, 31 Dec 2008 13:58:00 GMT

Justin Timberlake and cumulative advantage

http://community.brighton.ac.uk/jd29/weblog/39189.html

Full story at: http://jondron.net/cofind/frshowresource.php?tid=5325&resid=1386

Great study by Duncan Watts (of Six Degrees fame) that shows we are at least as influenced by what we perceive that other people think of a song as we are by the quality of that song.

It is all down to cumulative advantage – the early bird catches the work then the rich get richer while the poor get poorer. Quite aside from what this tells us about democracy and many other processes, this is at the heart of a fundamental problem that we need to overcome when designing social software, be it Google or Facebook or anything down below. Tricks such as parcellation, delay, limited sampling and so on can help, but as long as we have the channels to reinforce poor choices (be they Justin Timberlake, George W. Bush or Blackboard) then the problem will persist.
Created:Tue, 30 Dec 2008 18:01:12 GMT