Tuesday, July 31, 2012

Picture Post

Haven't posted many pictures lately. What can I say; scanning and processing takes time and I haven't been happy with my results lately. See if things improve going forward. Of course, anything becomes a chore this time of the year. It's 37° in Osaka today (which means it's over 40° locally where we live), and worse, nighttime temperatures never drop below 27°. Even sleeping is difficult, as you choose between stifling heat or clammy, miserable air conditioning.

McLaren is not just a formula 1 racing team; they apparently make sports cars as well. Or perhaps it's the other way around. Anyway, they've opened a showroom not far from where we live (and just a block or so from the new Ferrari showroom — is there one-upmanship going on?), and they allow photography in the shop as well.

A car like this — I could never afford to buy one. I don't have a license to drive it if I did. And I wouldn't have anywhere to drive it to its potential if I had one. My guess is, few of these cars sold in Osaka will ever go much faster than 100km/h or so in their entire lifetimes. They'll spend most of their life occasionally cruising down Midōsuji on sunny weekends. Many owners would probably be just as happy with an electric motor in the car, coupled with a good sound system to play V-12 engine sounds. Or happier; it'd be lighter, more reliable and get better red-light acceleration.

50 yen for a can of coffee. Deflation in action.

Instagrammy Bicycles
Instagram is a site and app that does "film shots" using phone cameras. I gave this shot an instagram-like treatment, just for fun. All shots in this post is from the same roll of cheap film and shot with an old, unserviced film camera; note how much violence I had to do to this particular real film picture to mimic the fake "film" effect of Instagram...

Covering The Beat
An incident in Shimanouchi, Osaka, and a gaggle (rabble? swarm? flight?) of journalists and television crews descended on the area. As always, stepladders were much in evidence.

News As It Happens
"So... I'll just stand here with the tripod then, shall I? Fine."

Canal Side
The canals in Osaka are generally not directly accessible. Only building owners and maintenance crews can get to the narrow strips of land between buildings and the canal itself. It's a shame; a canal promenade could be pretty pleasant in spring and autumn. Less so in mid-summer of course, when the canals turn rather fragrant.

Yokobori canal, which turns into Dōtonbori further downstream.

Friday, July 27, 2012

Olympic Post

Time for the (incredibly repressive) Olympics again, this time with an unusually suggestive logo. Events go live from early evening and on into dawn here in Japan, with commentary and highlights during the day. Starting tomorrow I will effectively be single for two weeks as my wife sits glued to the television. Time for some quiet reading and to catch up on my Japanese studies.

Ritsuko had some banking errands at the post office1 recently; the clerk made a mistake, however, and had to ask her back to the office. By way of apology they gave her these Olympic-themed piggybanks — piggybanks, plural, in a gift box. I think they are normally promotional gifts to new customers.


A set of three Olympic-themed gilded porcelain piggybanks. The shape is the old Japanese post box still used in some parts of the country. The Olympic connection is, in case you missed it, the medal colours of Gold, Silver and Bronze. Neat idea I guess, and they were fun, if tricky, to photograph.

But I'm not sure about the wisdom of passing these out widely; they're not exactly discreet, and many people have little need for any piggybank in the first place. We certainly don't, but it feels like a waste to just throw them away. I'll use them to collect 1, 5 and 10-yen coins for a while, just to see how fast they'll accumulate.

#1 Japan Post is one of the largest savings banks in the world; many — perhaps most — Japanese keep part of their savings there. Competitors used to complain that a part of a state ministry had such a large presence in the banking market, so the post office bank was spun off as a commercial subsidiary in 2008 or so. They're still just as big, of course, and as a private company they're now free to compete without the restrictions on marketing and services they had when part of the government. Be careful what you wish for is the message, I guess.

Wednesday, July 18, 2012

Robot Research Reports Reconsidered!

A note about a retracted nature paper (via Massimo Pigliucci) got me thinking:

Bentham Science accepts computer generated hoax paper for publication without realizing it.

A hoax paper is bad of course, and should be retracted. But computer generated? Computer-generated papers would be good in my view. Any idea just how much of people's time is spent drafting, editing and revising papers? Too much time. Time that you could use for doing your research, or presenting it, or reading up on other people's work.

Imagine a lab set-up with experimental apparatus and analysis system, all run pretty much automatically, once set up for a specific experiment. The system could be completely autonomous, such as a chemical reactor or a robot-assisted wet lab. Or it could be a semi-automatic set-up where a human does much of the actual work, but the system helps them record everything they do, collects all data from cameras, measuring devices and probes and so on. We already have this, pretty much, in many fields ranging from chemistry to physics to plant and animal biology to  behavioural research. We also all have more or less improvised analysis systems for crunching the data once we've gathered it.

That system effectively knows what goes in to the experiment, it knows all details of the experimental procedure — as it was supposed to happen as well as what actually did happen — and it has direct access both to the raw data and to your analysis of it.

It seems to me there is only a short step to use natural language processing to generate a fairly complete Methods and Results section for a paper without you having to enter a single word of your own. You'd simply select the data and the results you want to present (having to explicitly choose would reduce chance of inadvertent cherry-picking), and the system would churn out a complete text, formatted and adapted for the journal of your choice.

It won't rival the very best-written papers of course, but it doesn't have to; nobody expects deathless prose in a research paper. All it has to do is not be _bad_ and it will already be ahead of half the papers out there. Sure, it will be staid and formulaic, but in the context of a research paper that is a good thing. A benefit, not a drawback. You want to be creative and original with your research, not with your paper prose.

And the generated text will be more likely to be correct; the system knows what actually happened, what parameters where used when, in what order you did the analysis, and will not misremember, mix things up or forget. You would have to actively deselect particular experimental runs or data sets, and that decision would of course be on record for later in case someone asks whether your analysis is all aboveboard.

The introduction and the discussion would not be amenable to automation in the same way, at least not in the short to medium term (though the abstract is, I think, once the rest of the paper is done). What you could do is have the researcher draft the sections, perhaps in point form, then have the system interpret it and recast it in the language and formalisms best suited for the chosen journal. A secondary use would be automated revision and reformatting of a paper for a different journal if you're rejected from the first one.

Won't this take our jobs? no - it will let us spend more time working as researchers, and less as amateur editors. In the same way, lab automation simply frees up grad students and post-docs to work on original research, not as lab technicians. I do not fear a time when computers conduct experiments and write research reports. I await it, I embrace it and I can not see it happen fast enough.

Thursday, July 12, 2012

Should You Trust University Rankings?

I posted this on Google+ but I figure it fits here as well:

Times Higher Education university ranking has been released, with an interesting surprise. Alexandria University in Egypt — an good university but not remarkable in any way — ends up fourth in the world for research impact.

The reason: physicist, engineer and near-crackpot Mohamed El Naschie. He was founder and editor for a high impact-factor Elsevier journal, but was found to have published over three hundred of his own, unreviewed papers in the journal, and his rampant celf-citations both blew up the impact factor of the journal, and, apparently, the research ranking of his university.

So we have one sole researcher that has stuffed unreviewed papers into a journal he controls and cited his own work at every turn. That inflation of citations by a single researcher in one single department is enough to trump the massed outputs of major research universities. A major ranking category is thus easily accidentally gamed by a single person — I doubt El Naschie never even considered this possible side effect of his paper stuffing.

The irony is not lost on me, by the way, that this ranking is published by Thomson Reuters, who produce the impact factors in the first place. There is a kind of poetic justice in play here.

It shows just how much of a fraud the idea of impact factors is in the first place. The idea that the number of citations to papers could determine the quality of the journal they appear in was never a good one. It is crude and misleading; prone to gaming and manipulation by authors, journals and by Thomson Reuters themselves; and was only ever adopted because there was no better option available at the time. The impact factor is an idea whose time has well and truly passed.

Now, if a single researcher can manipulate an easy-to-measure ranking like this, imagine how easy it might be to manipulate the ranking in less easily measured areas such as the teaching environment. Consider how much these rankings mean in income for a university — your place on lists such as this one can make a major difference in your enrolment numbers, alumni donations and so on. And imagine how easily the system can be gamed if the university leadership — not just a lone researcher at a single department — decide that their university really deserves a better rank, and "adjusts" data to make it happen.

In other words, don't trust these rankings. Ignore them. There is a lot of rumbles about universities polishing their data, being selective with information, or blatantly making things up to improve their numbers for various rankings. These figures are being gamed; we just don't know to what extent.

And even if the rankings were accurate and unbiased, they are not informative for any one student or researcher. Even a small university is a sprawling place, with lots of variation in quality from department to department and from lab to lab. What you will get out of an education or a post-doc depends a lot less on the overall rank, and far more on the culture and atmosphere of the specific department and on the specific people you will have as teachers, co-workers and fellow students.

Tuesday, July 3, 2012

Political Clown Acts

Street Clown

You know how in some circus clown acts, the act sort of ends by mayhem breaking out, with everybody running madly around, honking horns, throwing pies and water buckets, dropping their pants, falling into tubs of custard, shooting confetti into the audience and generally just being as noisy and obnoxious as possible?

Think about such a scene. Fix the image in your mind. Got it? OK. Now you have quite an accurate image of the current Japanese political situation.

The DPJ is pushing through a consumption tax increase. Ozawa is — as always — taking his toys and leaving the party in a huff, taking his playmates with him into, presumably, another new party of his making. There's enough defectors to deny the government any ability to push through legislation without broad support from the opposition, but not enough for them to lose their majority outright and force an election.

..an election, by the way, that is currently impossible in practice, as the current election system has been ruled unconstitutional by the high court1, and the deadline for fixing it has passed. For an election to happen the houses have to pass a set of bills creating a new, constitutional election system. To do that, of course, they now need the support of the LDP and New Komeito in the opposition.
..an opposition whose sole idea for the past few years has been to vote against anything the government does in order to force a new election as soon as possible. And it's the opposition that has the most to lose from a new election system as it will remove power from small parties (such as New Komeito) and lessen the influence of rural districts (typical LDP strongholds). So they have to pass those bills for an election to take place, while they really, really do not want to. Oh, and both the DPJ and the LDP will have new leadership elections, adding a heady zest of internal power politics to the mix.

The passage of those bills and the results of the next election is really the only remaining thing that matters. At that point we'll see how Ozawa and the other small-fry parties will fare, whether Osaka mayor Hashimoto and Tokyo mayor Ishihara — who have been flirting in public for a while now — will make good of their threat to form a new national party2, or whether the LDP and DPJ will manage to hold together as unified parties.

Anything that happens up until that point is just clowns honking horns and throwing custard pies at each other.

#1 In short, the difference in voting power across the country was judged far too great; a single vote in some small rural districts can count for several times a vote in populous Tokyo or Osaka.

#2 Both come across as nationalist and populist strong-men; the main argument against a common party is that neither seems likely to accept anybody but themselves as the sole Big Guy at the top.