Journals of Negative Results

Via the INDUCTIVE mailing list, I learned of the Journal of Interesting Negative Results in Natural Language Processing and Machine Learning

It is becoming more and more obvious that the research community in general, and those who work NLP and ML in particular, are biased towards publishing successful ideas and experiments. Insofar as both our research areas focus on theories “proven” via empirical methods, we are sure to encounter ideas that fail at the experimental stage for unexpected, and often interesting, reasons. Much can be learned by analysing why some ideas, while intuitive and plausible, do not work. The importance of counter-examples for disproving conjectures is already well known. Negative results may point to interesting and important open problems. Knowing directions that lead to dead-ends in research can help others avoid replicating paths that take them nowhere. This might accelerate progress or even break through walls!

That’s healthy thinking, although the site/project/journal seems very new, not much up there yet. However it does have a page of links to other such journals, events, forums and articles in favour of documenting scientific failures. Listed in there is an upcoming AAA-08 Workshop, What Went Wrong and Why: Lessons from AI Research and Applications.

The workshop has its own Web site  with materials from an earlier 2006 event with intriguing abstracts from Douglas Lenat, John McCarthy and others.

The second workshop will continue our analysis of failures in research.  In addition to examining the links between failure and insight, we would like to determine if there is a hidden structure behind our tendency to make mistakes that can be utilized to provide guidance in research.

The forgotten poetry of Michel Thomas

(excerpts)

like that, that way
 it is like that
 it is not like that
 it is not possible for me that way
 it is not possible for me like that
 I'm sorry
 I'm sorry, but
 I'm sorry but it is not very comfortable for me that way
 I'm sorry but it is not acceptable for me that way
 it is very remarkable
 considerable
to go
 to come
 to go
 will you go eat with me
 where
 where do you want to go
 where do you want to go eat
 tonight, this evening
 it is for tonight
 it is for me
 it is for tonight
 where do you want to go eat tonight
 to have dinner
 where do you want to have dinner tonight?
 do you want to come have dinner with me tonight?

(further excerpts)

it is acceptable for you
 is it acceptable for you?
 isn't it acceptable for you?
why isn't it acceptable for you?
 like that, that way
 it is like that
 it is not like that
 it is not possible that way
 it is not acceptable for me that way
 why isn't it acceptable for you that way?
 I'm sorry
constant
 evident
 urgent
 it is very urgent
 I need it now, it is very urgent
 I have
 I have it
 I don't have it
 you have
what
 what do you have?
 what do you have for me now?
 I have
 I have it
 I don't have it
 you have
 what do you have?
 I have it
 you have
 you have it
 do you have it?
 you don't have it
 don't you have it?
 don't you have it for me now?
 why don't you have it for me now, because I need it now
 I want
 I want it
 I want it now
 I don't want it
 I want
 you want
 I want
you want
 what do you want?
 I want it
 I don't want it
 you want
 you want it
 do you want it?
 why don't you want it that way?
 to know
 I want to know
 I don't want to know
 I want to know why you don't have it for me now
 I can
 you can
 to do, to make
 to do like that
 what do you want to do now?
 to do, to make
 to eat
 something
 I want something
 I have something for you
 I want to eat
 I want to eat something now
 I have
 I am hungry
man
 I want to eat something now because I am hungry
 what do you want to eat?
 I am hungry
 you are hungry
 are you hungry?
 are you hungry? do you want to eat something now?
 why don't you want to eat?
necessary
 it is not necessary for me now because I don't need it
I'm sorry, but I don't have it and I don't want it because I don't need it now

WHY MIGHT CONNECTING WITH ZANDER JULES BE A GOOD IDEA?

Or: towards evidence-based ‘add a contact’ filtering…

This just in from LinkedIn:

Have a question? Zander Jules’s network will probably have an answer
You can use LinkedIn Answers to distribute your professional questions to Zander Jules and your extended network. You can get high-quality answers from experienced professionals.

Zander Jules requested to add you as a connection on LinkedIn:

Dan,

Dear
My name is Zander Jules a Banker and accountant with Bank Atlantique Cote Ivoire.I contacting u for a business transfer of a large sum of money from a dormant account. Though I know that a transaction of this magnitude will make any one apprehensive,
but I am assuring u all will be well at the end of the day.I am the personal accounts manager to Engr Frank Thompson, a National of ur country, who used to work with an oil servicing company here in Cote Ivoire. My client, his wife & their 3 children were involved in the ill fated Kenya Airways crash in the coasts of Abidjan in January 2000 in which all passengers on board died. Since then I have made several inquiries to ur embassy to locate any of my clients extended relatives but has been unsuccessful.After several attempts, I decided to trace his last name via internet,to see if I could locate any member of his
family hence I contacted u.Of particular interest is a huge deposit with our bank in our country,where the deceased has an account valued at about $16 million USD.They have issued me notice to provide the next of kin or our bank will declare the account unservisable and thereby send the funds to the bank treasury.Since I have been unsuccessful in locating the relatives for past 7 yrs now, I will seek ur consent to present you as the next of kin of the deceased since u have the same last names, so that the proceeds of this account valued at $16million USD can be paid to u and then u and I can share the money.All I require is your honest cooperation to enable us see this deal through. I guarantee that this will be executed under all legitimate arrangement that will protect you from any breach of the law. In your reply mail, I want you to give me your full names, address, D.O.B, tel& fax #.If you can handle this with me, reach me for more details.

Thanking u for ur coperation.
Regards,

I’m suprised we’ve not seen more of this, and sooner. Youtube contacts are pretty spammy, and twitter have also suffered. The other networks are relatively OK so far. But I don’t think they’re anything like as robust as they’ll need to get, particularly since a faked contact can get privileged access to personal details. Definitely an arms race…

Semantic Web Interest Group f2f meeting proposal (during week of Oct 20-24 2008)

I’ve started a thread on the Semantic Web Interest Group list, proposing that we meet during W3C’s Technical Plenary week this coming October. If you like the idea and plan to attend, please jump in and say so. If you have other ideas, please let us know them!

In the past we have handled this fairly informally, mixing short talks, themed discussion, inter-WG liaison, and lightning talks. This year I would like to theme any meeting around the practicalities of mainstream rollout: obstacles, issues, opportunities that arise as these technologies find their way into wider use. But this is a broad topic. What would you all like to discuss?

Comments and suggestions here or on the list please; although of course you can always ping me privately if needed.

Hope to see you in October…

Opening and closing like flowers (social platform roundupathon)

Closing some tabs…

Stephen Fry writing on ‘social network’ sites back in January (also in the Guardian):

…what an irony! For what is this much-trumpeted social networking but an escape back into that world of the closed online service of 15 or 20 years ago? Is it part of some deep human instinct that we take an organism as open and wild and free as the internet, and wish then to divide it into citadels, into closed-border republics and independent city states? The systole and diastole of history has us opening and closing like a flower: escaping our fortresses and enclosures into the open fields, and then building hedges, villages and cities in which to imprison ourselves again before repeating the process once more. The internet seems to be following this pattern.

How does this help us predict the Next Big Thing? That’s what everyone wants to know, if only because they want to make heaps of money from it. In 1999 Douglas Adams said: “Computer people are the last to guess what’s coming next. I mean, come on, they’re so astonished by the fact that the year 1999 is going to be followed by the year 2000 that it’s costing us billions to prepare for it.”

But let the rise of social networking alert you to the possibility that, even in the futuristic world of the net, the next big thing might just be a return to a made-over old thing.

McSweenys:

Dear Mr. Zuckerberg,

After checking many of the profiles on your website, I feel it is my duty to inform you that there are some serious errors present. [...]

Lest-we-forget. AOL search log privacy goofup from 2006:

No. 4417749 conducted hundreds of searches over a three-month period on topics ranging from “numb fingers” to “60 single men” to “dog that urinates on everything.”

And search by search, click by click, the identity of AOL user No. 4417749 became easier to discern. There are queries for “landscapers in Lilburn, Ga,” several people with the last name Arnold and “homes sold in shadow lake subdivision gwinnett county georgia.”

It did not take much investigating to follow that data trail to Thelma Arnold, a 62-year-old widow who lives in Lilburn, Ga., frequently researches her friends’ medical ailments and loves her three dogs. “Those are my searches,” she said, after a reporter read part of the list to her.

Time magazine punditising on iGoogle, Facebook and OpenSocial:

Google, which makes its money on a free and open Web, was not happy with the Facebook platform. That’s because what happens on Facebook stays on Facebook. Google would much prefer that you come out and play on its platform — the wide-open Web. Don’t stay behind Facebook’s closed doors! Hie thee to the Web and start searching for things. That’s how Google makes its money.

So, last fall, Google rallied all the other major social networks (MySpace, Bebo, Hi5 and so on) and announced a new initiative called OpenSocial. OpenSocial wants to be like Facebook’s platform, only much bigger: Widget makers can write applications for it and they can run anywhere — on MySpace, Bebo and Google’s own social network, Orkut, which is very big in Brazil.

Google’s platform could actually dwarf Facebook — if it ever gets off the ground.

Meanwhile on the widget and webapp security front, we have “BBC exposes Facebook flaw” (information about your buddies is accessible to apps you install; information about you is accessible to apps they install). Also see Thomas Roessler’s comments to my Nokiana post for links to a couple of great presentations he made on widget security. This includes a big oopsie with the Google Mail widget for MacOSX. Over in Ars Technica we learn that KDE 4.1 alpha 1 now has improved widget powers, including “preliminary support for SuperKaramba and Mac OS X Dashboard widgets“. Wonder if I can read my Gmail there…

As Stephen Fry says,  these things are “opening and closing like a flower”. The big hosted social sites have a certain oversimplifying retardedness about them. But the ability for code to go visit data (the widget/gadget model), is I think as valid as the opendata model where data flows around to visit code. I am optimistic that good things will come out of this ferment.

A few weeks ago I had the pleasure of meeting several of the Google OpenSocial crew in London. They took my grumbling about accessibility issues pretty well, and I hope to continue that conversation. Industry politics and punditry aside, I’m impressed with their professionalism and with the tie-in to an opensource implementation through Apache’s ShinDig project. The OpenSocial specs list is open to the public, where Cassie has just announced that “all 0.8 opensocial and gadgets spec changes have been resolved” (after a heroic slog through the issue list). I’m barely tracking the detail of discussion there, things are moving fast. There’s now a proposed REST API, for example; and I learned in London about plans for a formatting/templating system, which might be one mechanism for getting FOAF/RDF out of OpenSocial containers.

If OpenSocial continues to grow and gather opensource mindshare, it’s possible Facebook will throw some chunks of their platform over the wall (ie. “do an Adobe“). And it’ll probably be left to W3C to clean up the ensuring mess and fragmentation, but I guess that’s what they’re there for. Meanwhile there’s plenty yet to be figured out, … I think we’re in a pre-standards experimentation phase, regardless of how stable or mature we’re told these platforms are.

The fundamental tension here is that we want open data, open platforms, … for data and code to flow freely, but to protect the privacy, lives and blushes of those it describes. A tricky balance. Don’t let anyone tell you it’s easy, that we’ve got it figured out, or that all we need to do is “tear down the walls”.

Opening and closing like flowers…

Restarter martyr: sounds of Firefox 3b5

I use Firefox with a lot of tabs. I guess I’m a multi-tasker. Or I have surplus attention. Or they fixed enough memory leaks in Firefox 3  so that opening a new tab is almost cost-free. Until you restart your browser (is there a bug open for this? I couldn’t find one).

I’ve just made a quick movie from the unedited sound of Firefox 3 restarting this morning: Krafty Nacirema (multitasking mix).

Raw materials include: New Order’s ‘Krafty‘ and ‘Confusion‘; John S. Hall’s ‘America Kicks Ass‘ rant, and who knows what else. I liked how it all sounded, so here it is.

GraphPath

GraphPath

… a little-language for analysing graph-structured data, especially RDF. The syntax of GraphPath is reminiscent of Xpath. It has a python implementation that can be teamed up with your favourite python RDF API (e.g. Redland, rdflib, or your own API).

Anyone tried it? Apparently it has a rule/inference system too (backward chaining). I don’t quite see how namespace abbrevations are handled, but guess there must be some mechanism for registering them in the API. Could be a candidate for transliteration into Ruby?