David's World

Cognitive Modeling and Computational Linguistics

2012-01-25T02:28:08Z

This workshop provides a venue for work in computational psycholinguistics: we invite a broad spectrum of work in the cognitive science of language, at all levels of analysis from sounds to discourse. It will be held at NAACL-HLT'2012 in Montreal.

Find the full Call for Papers here.

Full-screen mode in Mac OS X "Lion": a solution to the DVD watching problem

2011-12-06T22:07:56Z

Lion's concept of reserving a separate space for full-screen windows is great in principle. However, this doesn't work so well on multiple-display setups. One example is DVD Player not being able to play a movie on a secondary display. Even when the external display, such as a projector, is set up as the main display, the user won't be able to do work while projecting the movie. One idea for a friendly and intuitive fullscreen concept would be to modify the function and appearance of the fullscreen button (top right of window) if the window is on a secondary display. In full-screen mode, the window will then behave as on pre-Lion systems: it will not occupy its own space, and the primary display is retained for regular use. In this mode, the fullscreen window on the secondary display is the topmost layer, and all other displays behave as if the secondary display was not part of the display set. In particular, switching screens will not change the secondary display. (Internally, when the user switches screens, the full-screen window is moved automatically to the new screen and kept as full-screen, top-most layer, so it will appear static. The animation for screen-switching will have to adjusted accordingly.) This mode of interaction seems conceptually intuitive, even though it introduces a second kind of fullscreen mode. I think it matches the use cases that people have quite well.

Pilot's Logbook

2011-09-16T18:45:54Z

I've uploaded the first version of my Excel spreadsheet for pilots, which gives you an electronic log book. While geared towards glider pilots, it should also work fine for other General Aviation pilots (who may need to add a place of landing and a large list of airports). The log book prints statistics about currency as pilot-in-command, in single/multi-seaters, different launch types and high performance sailplanes, and it minimizes the work needed to make entries. It's free, licensed under the GNU General Public License, and available here. Please send me useful changes, ideally via Github. I can't promise to reply to support requests due to my general workload.

The new rockstars

2011-01-13T22:44:45Z

iPhone 3G with iOS 4.x and jailbreak: Fix for crashes and a battery drain issue.

2011-01-11T15:27:26Z

Plenty of trouble with my jailbroken iPhone 3G, which I run on T-Mobile in the US. Apple offers its iPhone bundled with an AT&T contract, which keeps the initial payment down (attractive for delusioned customers) and the total cost of ownership up (good for Apple and AT&T.)

To use an iPhone with any GSM network (such as AT&:T or T-Mobile in the US, and virtually any provider in Europe), it needs to be "jailbroken" and "unlocked". The process is simple: download a program called "PwnageTool" from a website, run it over the iOS operating system that was downloaded with iTunes (.ipsw), then install the operating system on the iPhone with iTunes. Once the firmware update has been installed, start the new Cydia app on your phone, find the Ultrasn0w application and install. That's it: your iPhone will now accept GSM SIM cards from all carriers and in all countries. PwnageTool only runs on Macs. If you are still on Windows, have a look at a program called Redsn0w.

The real know-how is in the details, and the iPhone-Dev team makes you read a lot of text before getting a corse idea what to look for. On top of that, crashes and battery drains have made my life difficult over the last few months. Here are the solutions I found. The apply to the iPhone 3G, but not to later iPhones (3GS, iPhone 4).

Installing iOS 4.0, 4.1 and 4.2 on iPhone - slowness, crash and reboot issues

Lots of trouble. First the incredible slowness IN IOS 4.0. The solution was to upgrade to 4.1 when it finally came out. This wasn't as slow, but buggy in conjunction with the iPhone-Dev team's Jailbreak. Crashes galore all complete with a reboot that took many minutes, just when you're urgently looking for directions. A clean slate supposedly helped: restore the iPhone via iTunes and a jailbroken operating system file but one had to avoid restoring one's data and apps from a backup in iTunes. The manual restore is easier done than it sounds: ITunes synchronizes all data from the Address Book and Mail.app. Of course, if one keeps mail accounts or phone numbers in the phone without syncing them back to a computer, the install process will be much more involved. This solved the crashes for me.

I upgraded my phone to iOS 4.2.1, with the iPhone-dev jailbreak and the ultrasnow unlock. The trick here is to make sure that the phone's baseband version is compatible. Then, no baseband upgrade should be performed. Pwnagetool will ask you whether to upgrade the baseband.

Battery drain issue with iOS 4.x

The ordeal wasn't over for me, though. Now I dealt with a phone that drained its battery At an impressive rate - it would not last an afternoon. Online pundits have suggested many solutions to the problem, but the correct one appears to be that the iPhone keeps checking for "push certificates" via a Wifi or EDGE or 3G data connection. These certificates fail to validate, and the checking goes on. The Data Usage statistic (in Settings -> General) demonstrates these transfers.

One solution is to use a new version of Redsn0w for the jailbreak. If using Ultrasn0w, install a further Cydia package called "Push Doctor". This has fixed the battery drain issue for me and others. Important note: uninstall OpenSSH afterwards, or at least change the iPhone's root and mobile user passwords.

Aquamacs, Lisp and more outdoorsy fall fun

2010-12-24T15:15:28Z

Some Aquamacs and Lisp-related pointers, and some non-programming flying fun pictures taken this fall.

A good use of Lisp on Deep Space One, which was not a Star Trek pilot, but a fascinating NASA program testing modern technologies deemed too high-risk to use on an expensive mission vehicle. Deep Space 1 ran a Remote Agent programmed in Lisp! (Harlequin Common Lisp, to be precise, which is now called LispWorks.)
From a blog: And then, it happened. I came across Aquamacs...
Getting ready for Aquamacs 2.2: Lots of bug fixes and improvements in Aquamacs before the year ends. Test-drive the 2.x and 3.x builds here.
Some aviation pictures taken on the last warm weekend of the year at KCBE (Greater Cumberland Regional Airport): a Cessna 152 landing, me landing behind a Citation, and a (Schweizer?) glider waiting to launch.

Happy holidays!

Link Soup: October. What's cooking (in the world)

2010-10-11T17:53:01Z

Interesting reads and sights that didn't get broad attention in the mainstream press: Durrell's vontsira - a novel carnivore species discovered on Madagascar. And a cute one! (Big pictures here.) The Guardian about Jon Stweart, who misses an imaginary Hero-Obama. Ingrid Betancourt about her time in captivity. People (and politicians) do what's best for them - even if that isn't immediately obvious: Elizabeth Pisani's TED talk. NELL - the never-ending learning system learns about semantics. A 15-minute interview with CMU Machine-Learning professor Tom Mitchell, whose team created this.

Link Soup. (What's cooking in David's life...)

2010-09-15T22:15:17Z

Lack of blogging does not represent lack of activity in a blogger's life. The opposite holds true. Yet, I'd like to share a few pointers to interesting or fun things that I've seen this week. My trusty Facebook friends have seen those first, but why keep things under wraps?

How the Taliban recruit suicide bombers (and how life in Afghan villages escapes the imagination of US politicians and militaries): Sharmeen Obaid at TED.

Carnegie Mellon's Snake Robots.

My pictures: Gliding over Willamette Valley, Oregon. A glider crash in the UK that ended better than you'd think from looking at this amazing series of photos.

Inside the Secret World of Trader Joe's.

Aquamacs 2.1 was released in August, providing improved stability and speed. And guess what - Aquamacs 2.2 will be even better.

Playing the Lemonade Game: Metacognition in Game Simulations.

2010-03-26T15:32:59Z

A few details about our work on the Lemonade Stand Game leading up to our entry in the Lemonade competition at the end of 2009 can be found in our latest paper, presented at this weeks Behavior Representation in Modeling & Simulation conference. The competition was a tournament between agents, playing out many strategies that all tries to succeed to win lemonade market share on the beaches of a fantasy island. Abstract: The Lemonade Game is a three-player game in which players have to pick locations on a circular board, which are as far away as possible from those chosen independently by other players. Players may observe other player's moves and infer their strategies. The game was studied using a competition of cognitively motivated agents, which inherit properties of adaptivity and stochasticity from human memory and decision-making, and simplistic, yet effective agents implementing fixed strategies. We argue that metacognition is the unique attribute that allows sophisticated agents to adapt to unforeseen conditions, cooperators and competitors. D. Reitter, I. Juvina, A. Stocco, and C. Lebiere. Resistance is futile: Winning lemonade market share through metacognitive reasoning in a three-agent cooperative game. In Proceedings of the 19th Behavior Representation in Modeling & Simulation (BRIMS), Charleston, SC, 2010. Metacognition was also explored in my cognitive model submitted to the Dynamic Stocks&Flows Competition last year. There, the model did just what I think human participants in a psychological experiment did: try different strategies to solve a problem, and pick the one that performed best most recently. Weighing recency and frequency of experiences or pieces of knowledge is a common theme in such models; more research is needed to find out how exactly people combine their experiences to make (mostly) good decisions. The model won the first prize in the challenge. Still, the model does not explain enough of the variance observed in the empirical data. A look back to determine what we have learned from the DSF challenge was presented in a panel discussion at this year's Behavior Representation in Modeling & Simulation conference.

US Airways

2010-01-08T14:17:10Z

This winter's itinerary: Pittburgh-Charlotte,NC-Frankfurt,Germany, and Munich,Philly-PA,Pittsburgh, at a whopping $900+ price tag. Results: Not flying with US Airways again if I can avoid it. My First flight was cancelled due to some software problem with an onboard computer on their Embraer jet, after we've been on board for an hour. Rerouted - next flight 4 hours late due to a failed hydraulic pump. Then a fully booked overnight flight in the wrong kind of seat. Arrival: 7 hours late on the other side of the Atlantic in the end, no compensation from US (only EU airlines are required to reimburse/make happy their customers). I've learned a lesson about redundancy: when you fly aircraft that have 12 hydraulic pumps (not all redundant), you don't just increase safety while in the air - you also increase the likelihood that the aircraft is going to be grounded because one of them fails. (Note to self: compile list of airlines that charge for infants.) Return flight: East Coast in the winter chaos; plane running one hour late as it is being de-iced. Is ice in January such a surprise that this couldn't have been anticipated? No more US Airways flying for me. It does not suffice that their cabin staff has always been super-friendly, and that they have a pilot who can fly very big gliders on staff. The getting-me-there-on-time bit has to work. Rant finished.

Building small, big and fast aircraft: The power of startups

2009-12-17T17:18:01Z

Just saw a video (see below) of the newly-built SpaceShipTwo aircraft, a plane that will take a small number of super well-off to the end of the universe as we know it, or just above the atmosphere, to be precise. It's a fascinating design, but what's much more fascinating is the enterpreneurship story behind this. Since 2004, two small companies managed to develop two spacecraft and two carrier aircraft (jets); the first combination winning the 10M$ X-Prize in 2005, the second in the process of achieving FAA approval. Second, take the gliders (sailplanes) that I enjoy flying. All of them are developed and built at affordable prizes by small companies such as the German companies Grob, Schempp-Hirth, or Schleicher. They are built to withstand much higher G-forces than those airliners and reach amazing efficiency and high speeds. (Great powered aircraft are similarly built by small and medium-size aviation companies such as the American Cirrus.) Those gliders have been made from composite materials (fiberglass+gel coat) since the late 70's. I'm just mentioning this because I'm now introducing the herculanean efforts building the Boeing 787, the first composite airliner, taking the dedication of thousands of engineers and mechanics over a decade to build. The Airbus A380, the world's largest airliner, wasn't different. Of course, the comparison is ridiculous, even in the case of the pressurized, jet-propelled space craft+carrier planes that fly faster and higher than airliners, and that are built to comparable safety standards. Is the technical challenge that different when scaling up a pressurized jet from a dozen to a few hundred passengers? It seems clear: aircraft development doesn't scale well. I bet that very similar cases can be made for software development - compare the behemoth that is Windows (maintained by thousands of programmers at Micrsoft) to the Mac operating system that was initially developed by far fewer people at NeXT. Why is that so? It's a fascinating and unanswered question from the point of view of research into emerging collaboration in teams and of communication networks to explain the apparent loss in efficiency.

Are we better off without religion?

2009-12-09T02:53:45Z

Christmas is coming up, and just in time the Guardian goes over some recent research in the social sciences asking the question: are we better off without religion? Sociologist Gregory Paul's view is that religion is a sign of dysfunctional societies. Generally, societies with higher "popular religiosity" also fare worse when it comes to measures of community success, such as the number of jailed people, sexually transmitted diseases, corruption, and the like (Paul, 2005, 2009). The causality of this is not clear, but Paul argues that religiosity is the result rather than the cause. One should note that such correlations depend on a subjective, albeit wide-ranging Successful Societies Scale: if socio-economic factors would be weighted heavily, the most-religious country (US) would end up close to the the least-zealous societies (Sweden, Japan). One may wonder how much of the originally intended "good" is left in the world's religions - the good that stabilized societies through more or less arbitrary rule systems. The good that kept people healthy ("no pork!") and made communities stronger ("Love thy neighbor!"). The good that allowed some preachers and some movements out of many to evolve and develop into world religions. Note Pope Benedict II's (Joseph Ratzinger's) views on liturgy, which argue that the "new" forms have made church-goers self-indulgent and ignorant of views from outside (the priest turning to the congregation, preaching in their language - his conclusion: back to the Latin mass, is, of course, ridiculous). Christmas traditions are a good example of a combination of religious and pagan traditions that have lost their meanings. I for one am glad to avoid Santa Clauses, Christmas shopping in stores, and I greatly enjoyed my department's "non-denominational Holiday party". Paul, Gregory S., Cross-national correlations of quantifiable societal health with popular religiosity and secularism in the prosperous democracies: a first look. Journal of Religion & Society. 7, 2005. Paul, Gregory S., The Chronic Dependence of Popular Religiosity upon Dysfunctional Psychosociological Conditions. Evolutionary Psychology 7(3). 2009.

Improvised Explosives in Iraq: Of makers, disarmers, and the dead.

2009-08-16T00:26:31Z

The Hurt Locker is a film about a US bomb squad dying and cheating death in the heat and dust of Iraq. It is fiction, shot and edited in documentary style. Creepy Realism. Coming out of the movie I am still wondering: why did the US decide to send soldiers to hell, and how could the US-UK coalition commit the arson that removed a dictatorship to create the Iraqi hell in the first place? Questions, not new. The Hurt Locker is out in US theaters now. It's worth seeing. This week, the New York Times ran a remarkable piece titled How Baida Wanted to Die about a Iraqi suicide bomber who did not manage to explode her bomb. It is wide-spread religion that brainwashed a young woman. But the horrors of war and her marriage triggered her decision to join those who carry their belt stuffed with explosives to a busy market. A police officer told the reporter before her interview with the young woman: "You're going to like Baida".

How people find a common language

2009-06-08T18:44:15Z

Humans carrying out a task together find a common language, sooner or later. Even communities of speakers do this, and in fact this how they may come up with natural languages over time. At Carnegie Mellon University, we have now simulated this process in a cognitive computer model, tracing the steps that humans take to learn the elements of a graphical language, to forget them, to re-learn, and to arrive at a common language in a small community. Psychological research provides a pretty precise picture of how human memory works. In this paper to be presented at the International Conference on Cognitive Modeling, we describe the combination of two worlds: multi-agent and evolutionary simulation on the one hand, and psychologically validated models of memory and cognition to simulate the evolution of a domain language. The model explains the empirical data pretty well, but also makes a prediction: that human communication between two partners needs to go both ways in order for us to learn and to converge. David Reitter and Christian Lebiere. Towards explaining the evolution of domain languages with cognitive simulation. In: Proceedings of the 9th International Conference on Cognitive Modeling (ICCM), Manchester, UK, 2009.

]]> Abstract

We simulate the evolution of a domain language in small speaker communities. Data from experiments (Garrod et al., 2007; Fay et al., 2008) show that human communicators can evolve graphical languages quickly in a constrained task (Pictionary), and that communities converge towards a common language even in the absence of feedback about the success of each communication. We postulate that simulations of such horizontal evolution have to take into account properties of human memory (cue-based retrieval, learning, decay). We implement a model that can draw abstract concepts through sets of non-abstract, related concepts, and recognize such drawings. The knowledge base is a network with association strengths randomly sampled from a natural distribution found in a text corpus; it is a mixture of knowledge shared between agents and individual knowledge. In three experiments, we show that the agent communities converge, but that initial convergence is stronger when communities are structured so that the same pairs of agents interact throughout. Convergence is weaker in communities when agents do not swap roles (between recognizing and drawing), predicting the necessity of bi-directional communication in domain language evolution. Average and ultimate recognition performance depends on how much of the knowledge agents share initially.

Recession Era Victory Garden in Oakland, Pittsburgh

2009-06-05T21:51:02Z

Here's a great little Recession Era Victory Garden providing seedlings for everyone who likes to take some and plant them in their yard. (Found on Forbes Ave, Oakland, Pittsburgh.)