Fortnightly Mailing

Crawling into the skin of a user....

A few weeks there was "climbing inside your users' heads and seeing your web site through their eyes". In Beyond Search, a 22/11/2006 article in Information World Review, by David Tebbutt, we have the equally extreme “a good information professional should be able to crawl into the skin of the user and understand their thought processes", in this case a quote from Susan Feldman, from the market intelligence firm IDC, in support of the argument that . The focus of Tebbutt's article is the shake-up that is happening in the commercial search world, stemming from the way in which vast tracts of content are now being routinely indexed, with storage costs no longer really relevant, and the increasing sophistication of the software systems that determine how search terms are turned into search results.

Where does this leave the information professional? According to Tebbutt:

"it means that if your department is not already integrated with the organisation, it soon will be. Whether the initiative comes from you, from IT or from somewhere else, the fact is that your world, the business world, and the IT worlds will merge."

On a related issue, and straying into territory that I observe rather than in which I have professional expertise, here is an interesting report of a discussion between Tim Berners-Lee ("father" of the Web) and Peter Norvig (Director of Research at Google) at the July 2006 American Association for Artificial Intelligence conference. This  highlights the "divide" between people who want information and the links between it to be curated, with thought given to how both are described, and people who think that as more and more of the content "out there" is produced by amateurs, never going near an information or ICT professional, the only way forward is for search tools to extract meaning from the content without reliance on how that content has been categorised.

One laptop per child - further information and progress

David Weinberger has two 28/11/2006 posts [1], [2], including pictures, which which work as "latest news" about the One Laptop pre Child laptop, ased on a presentation by SJ Klein at Harvard's the Berkman Centre.

Previous posts which may also be of interest:

• What would you install on one laptop per child? 17 October 2006 Guest Contribution by Steve Ryan from talk at LSE by Jonathan Zittrain;
• The "One Laptop Per Child" wiki. 1 August 2006 posting which includes a video of a working prototype of the laptop.

Learnometer: "dashboard" measures of investment performance in learning

Stephen Heppell has a Microsoft-sponsored project to create a Learnometer, i.e. some "dashboard" measures of investment performance in learning:

"All round the world countries are investing significant and increasing amounts in education, particularly in ICT and in buildings. At the same time clear emergent learning trends, effective and consensual, are re-defining learning in the 21st century. The two fundamental questions are:

• is our investment in learning taking us in the right direction, towards 21st century learning? and
• if it is, how do we know how effective that investment has been - what should improve?"

Referring back to the post below about Andrew Pinder's call for education to be organised industrially, it would be interesting to get Pinder and Heppell into public conversation on these issues.

Organise industrially. The Chair of Becta's vision of e-learning. Webcast of an 80 minute seminar.

This is an 80-minute recording of a seminar held on 16 October 2006 at the Oxford Internet Institute (OII) addressed by Andrew Pinder, Chairman of the British Educational Communications and Technology Agency (Becta), the Government agency responsible for ICT in the whole of English public sector education with the exception of higher education.

Pinder spent a career in the financial services sector, and was for a time the Government's e-envoy. The seminar, which is chaired by OII's Bill Dutton, opens with Pinder's scathing (and glib?) assessment of why, despite billions of ringfenced Government spending, 85% of schools are getting next to no improvements in their effectiveness from the spending. Pinder's line - illustrated with a sideswipe reference to electronic whiteboards as a technological means of maintaining the status quo - is that unless education is organised industrially (that is, if it continues to be run as a cottage industry, with professional interests dominating) the money will continue to be wasted.  Becta is increasingly influential, so for English readers with an interest in public sector education, make a point of watching this, and stick with it to the final 20 minutes of open discussion. If I worked for Becta, I'd be nervous.

Update, 26/11/2006. David Smith, a public school ICT manager, provided an interesting reflection on Pinder's session (25/10/2006); Donald Clark posted a longish piece about the session (24/11/2006), in a similar vein to his earlier comment on this piece below; Julian Todd, an Open Source software programmer in Computer Aided Manufacture, writes in a hostile vein (20/11/2006), picking up on Pinder's comment about the individuals in schools whom he asserts tend to be responsible for procurement in schools  ("Typically they would be people who have a real passion about Open Source — as if open source is any different to any other software — it’s just the pricing structure is different, that’s all. But they have a passion. It’s a religion, it’s a real belief, and again they have a belief about bits of technology that are going to change things. What they don’t do, however, is organize things properly."). (Todd has also been busy with a Freedom of Information request concerning Becta's framework contract for the supply of hardware to schools.) Meanwhile, Becta is recruiting 3 new members to its Board of Directors [100 kB PDF].

Founders of iRows, the multi-user web-based spreadsheet, recruited by Google

Google has been in the news for its acquisition of YouTube. But it is also consolidating its grip on the web-based application market. In March I described iRows, a web-based multi-user spreadsheet, speculating that Google (which had just bought Writely, a web-based multi-user wordprocessor) might be interested in iRows, an Israeli start-up. Earlier this month, iRows stopped taking registrations, and will stop operating at the end of 2006. The two founders of iRows, Yoah Bar-david and Itai Raz, have been recruited by Google, which already has its own web-based spreadsheet system, sitting alongside its (ex-Writely) wordprocessor. Meanwhile, Google has bought JotSpot, a sophisticated wiki application, which has been used by the TUC for its on-line course development and management manual. When JotSpot starts taking registrations again - whether as JotSpot, or as a Google-branded application (one assumes the latter), it will be a free service. 

(If I had to predict what's next, I would say that Google will acquire a web-based audio and/or video-conferencing service.)

Machine translation. A 2005 article; and the November 2006 (and June 2008) NIST results.

NIST data set		BLEU-4
Site ID	Language	Overall	Newswire	Newsgroup	Broadcast News
google	Arabic	0.4569	0.5060	0.3727	0.4076
google	Chinese	0.3615	0.3725	0.2926	0.3859

GALE data set		BLEU-4
Site ID	Language	Overall	Newswire	Newsgroup	Broadcast News	Broadcast Conversation
google	Arabic	0.2024	0.2820	0.1359	0.1932	0.1925
google	Chinese	0.1576	0.2086	0.1454	0.1532	0.1300

Summary score table from NIST "Unlimited Plus Data" track

I included something about machine translation in the (pre- web log) Fortnightly Mailing Number 54. This June 2005 article by Gregory Lamb in the Christian Science Monitor is jargon-free, and explains the difference between the two main approaches to machine translation: rules-based - as developed by Systrans, and still used by Google; or statistically-based, as being developed by Google. (4/12/2006 - see also Not Lost in Translation, from the MIT Techology Review, by Stephen Ornes.)

As to the effectiveness of statistically-based methods, Google's system continues to score considerably better overall than the competition (who may or may not be using statistically-based methods), in both Arabic to English and Chinese to English, with the margin rather bigger for Arabic to English than for Chinese to English, in the US Government's NIST Translation Evaluations of machine translations of different genres of text (Newswire, Newsgroup, Broadcast News, Broadcast Conversation). Bear in mind however that a score of 0.503 out of a maximum of 1 (the best score achieved by Google for the translation from Arabic to English of Newswire genre, with  scores for Chinese to English consistently worse) does not mean that the absolute quality of the translation was particularly high.

It is also worth noting, although the test regime was different, so this observation needs taking with a pinch of salt, that Google's results do not seem to have improved much on what was achieved in the equivalent 2005 NIST tests.  (Here, for reference are the 2008 NIST results, which I have not had time to analyse.)

Updated 4/12/2006 and 1/1/2009

Interesting discussion about the wisdom or idiocy of crowds

Clive Shepherd starts a discussion, triggered by David Freedman's The Idiocy of Crowds - which challenges the "conventional wisdom" that groups take better decisions than individuals - to which (22/11/2006) Donald Clark and Mark Berthelemy respond. To varying degrees they are sceptical about the value of "small group work" in learning. DC "The most effective learning takes place on your own or one-to-one". MB "... the activities that rely most on the group are the ones that I feel less ownership of, and have far less relevance to my situation. The group activities are also far more work - mainly to deal with group dynamics issues rather than the content of the activity itself."

What do I think about this?

Usually my understanding (i.e. what I learn) develops if i) I have to express myself about the issue - verbally or in writing; ii) what I say or write about the issue is challenged by others. So on the face of it, "small group work", whether face-to-face, or on-line, should be suitable - for some things at least, depending of course on task-design, group-composition, and the useability of the technology. And in some contexts, learners in a small group can give each other the personally relevant and motivationally effective formative feedback that helps develop their learning (for more on this see Inside the Black Box), more cost-effectively than can a teacher - assuming a good one is available, and probably more effectively than can a piece of interactive software.  But small group work on how to write an Excel function, the chemical pathways in photosynthesis, or the mathematics of font-design? Other than as a means of developing confidence in expressing concepts, probably not.

Moodle-based provision wins Times Higher Award for "Most Imaginative Use Of Distance Learning"

Behind the par-for-the-course language in this 16/11/2006 press release from the University of Derby lies the fact that Derby's Learning Through Work provision, run for/in collaboration with Ufi/learndirect - one of the world's largest scale providers of online learning - is implemented using the Open Source VLE Moodle.

Simple email and asynchronous conferencing still have a lot going for them

In the mid-late 1990s I was quite closely involved in starting off some online distance learning courses which have stood the test of time, including the South Yorkshire FE Consortium's Learning To Teach On-Line course (LeTTOL) and, the Sheffield College's GCSE English Online; and I have written previously about the range of online courses that the college now runs.

These courses are characterised by:

• reliance on email and asynchronous text-based conferencing for most of the learner/learner and learner/tutor interaction;
• modestly designed, simple, web-based learning materials,  developed at a relatively low cost per learner hour of material;
• course development, maintenance, and delivery in the hands of a team of tutors;
• more-or-less all team members required to complete the LeTTOL course.

The college's courses are outstandingly successful in two senses. They:

• reach learners who would not normally manage to start, let alone stick with, a college course;
• have unprecendently high examination pass-rates.

Of course you can read too much into awards. Plenty of organisations do not bother to enter for them, which does not mean they have nothing good to show; and there are examples of awards being won for suspect activities. Nevertheless, I think it is significant that on 9/11/2006, in a bid organised by the South Yorkshire eLearning Programme (eSY), against competition mainly from the corporate sector, the English Online team and eSY won the national 2006 Award for the "best example of supporting learners on-line", organised by the magazine e-learning age. Currently I am encouraging Julie Hooper, who leads the team at The Sheffield College, has agreed to write a Guest Contribution for a future issue of Fortnightly Mailing. (See also Donald Clark's drinking champagne from shoes, a report from the awards ceremony.)

Web 2.0: the key to wealth? Zoho, Lala, Guba

Nice piece from 20/10/2006 by Ivor Tossell in the technology section of The Globe and Mail poking fun at the Web 2.0 world, and drawing a parallel with the situation during the first dot.com boom. Excerpt:

"Broadly speaking, the term refers to sites that fill their pages with user-submitted content, like blogs or photos or comments, and play up the Web's social-networking potential. Enthusiasts heralded Web 2.0 as a breakthrough in digital democracy, letting everyday users create media content. It also happened to be a heck of a lot cheaper to let users generate content (or, frequently, steal it) than to pay people to write the stuff.

Web 2.0 sites have gained a degree of legitimacy, thanks especially to a few high-profile success stories, like the formerly Canadian photo-sharing site Flickr.com, which Yahoo bought last year. And in their success, they've spawned an entirely entertaining cult of conformity, with cookie-cutter start-ups emerging weekly, laden with funny names and purposes that are all more of the same. Yet they keep popping up, and they keep getting funded. This, my friends, is where the money is -- if you toe the line."

Via The Top Ten Lies of Web 2.0, from where the image below is culled, and Stephen Downes.

27/4/2007. See also Charlie O'Donnell's Top Ten Reasons Why Web 2.0 Sucks.