Wednesday, December 29, 2010

Back to Fundamentals: What IS Interdisciplinary Computing?

Yesterday over coffee I had one of those really stimulating conversations with an acquaintance who works in a large software company. The topic was interdisciplinary computing. With hindsight, I now see this as a prelude to a larger conversation on the same topic ("what IS interdisciplinary computing?") that I expect to have late next week with a larger group of people from academia and industry.

It wasn't long into our meeting before I was reminded of the very fundamental excitement and challenges that start with just trying to define the term and jump off from there.

Some questions we enthusiastically tangled with:

What IS interdisciplinary computing? Is it... computer science embedded in other fields (e.g. earthquake simulation) where the people involved may have CS or other backgrounds (e.g. earth science or physics)?

If yes, does this count as "applied computing" and not "true computer science" (whatever that means, and we followed that conversational path for a while).

Is interdisciplinary computing the embedding of computational thinking in another field? (please define computational thinking....everyone is using it these days, but a commonly accepted definition is elusive)

Is interdisciplinary computing the equality of a merge between two fields - and within the educational system does this mean within a university computing department...or a cooperative endeavor b/w two departments (communications/digital media and computing for example), or a formal merge of two or more departments?

How do you successfully connect with the arts and humanities? When you move away from other STEM disciplines the challenge in establishing effective communication and trust magnifies, but so do the wonderful possibilities. We ran with this topic for a while.

What about K-12? This is a topic we did not delve into too deeply this time, but it looms large.

What does all of the above have to say about the content of a degree program in computing? The never ending question about the first year of coursework ... most of you are probably quite familiar with this line of discussion. Programming, not programming, programming with a new twist or theme or application, ... how to show the ubiquity of computing in that first year in an engaging way, which is the time period when we lose most of the students we lose?

What does industry want in its graduates in terms of interdisciplinary computing skills - it depends who you ask. In some meetings of the ACM Education Council a while back, we invited panels of industry speakers to talk about what they are looking for in graduates. The answers were all over the map and were influenced by such factors as size (large corporation vs. smaller company), corporate and divisional focus (product development, research, technical marketing etc).

If ideally, a liberal education is supposed to teach students to think critically, how does one's interpretation of interdisciplinary computing, the competing demands of those who might hire your graduates, and enrollment concerns tie in with philosophical stance on a computing education?

Speaking of philosophical stance...the question came up at one point: should computing departments even exist? Ooooo. So many holes to fall into in tackling that question.

If what form? If no, what should happen instead?

We considered these and other questions until the remains of my latte were cold and the foam solidifying in my mug. I suspect we could have kept talking had each of us not had other things to take care of yesterday.

Periodically being brought back to the fundamental questions of "what is" and "what does it imply" are very important when considering implementing large scale change - as many of us are either doing or considering doing.

Thursday, December 23, 2010

Charles Babbage the Interactive Auto GPS

Oddly enough, just 2 posts after I wrote about the development of s/w to recognize emotions that started out with a discussion of (imo) annoying auto GPSs, I find a short article and video about a British researcher named Peter Robinson who is developing a quite sophisticated auto GPS system (full story here). The video starts out by echoing my sentiment of how annoying and sometimes inaccurate a car GPS can be, and he actually tosses one in the back seat! I LOVED that moment. And he does it with such poise and non-violence. Just flips it behind him.

Then he discusses why it would be useful to have an interactive conversation with a computer - the computer can read your facial and bodily expressions, determine your emotional state and respond with its own emotion inflected conversation. That is both technically interesting and potentially useful in many areas of society.  I bet you can think of a few areas where this technology could be used productively?

But then he proceeds to discuss how he is testing the system on an auto GPS. Why oh why the fascination with the auto GPS??? Ok I accede to this as a useful example of holding a conversation, where important decisions have to be made on short notice, and there are unpredictable behaviors and circumstances to be dealt with. The creepy part is when Robinson pulls out a custom made head of Charles Babbage (looks a bit rubbery) and props it up at head level in the passenger seat of a car in a driving simulator. The two, Robinson and Babbage, carry on a very polite conversation about road conditions. Robinson ends by saying "Charles, I think this is the beginning of a beautiful relationship".

I have mixed feelings about the idea of driving down the road talking to a life size rubber head with wires coming out the back of it.

Technical information is included in the short video, and there is an interesting bit about how it all works. There are definitely compelling issues of emotional intelligence, neuroscience, psychology, and physiology here to explore. Robinson sees this development as the future of how we will interact with computers. Maybe AI has a not too distant future in realistically simulating life? That would be great progress indeed after years of inching along.

But the rubber head... I don't know about the rubber head. If it made mistakes and the driver got irritated, and tossed it in the backseat, would it start complaining from its disembodied self face down on the seat?

Sunday, December 19, 2010

Software and Art Authentication

I spent a few days up in the mountains unplugged. However, technology has an interesting way of creeping in even with the cellphone turned off (reception is unreliable anyway) and the laptop stowed. And I'm not going to talk about the visiting electrician who suggested that I test a 9V battery by licking it (he found this quite amusing, but I declined).

First, I was reading one of my alumni magazines and there was a story about a physicist who used software to examine highly controversial paintings that might or might not have been newly discovered Jackson Pollack works (full story here). The physicist, Richard Taylor, uses software to discover the presence of fractal patterns. Apparently, according to his research, true Pollack paintings are fractal filled. Benoit Mandelbrot agreed. Taylor also determined that fractal patterns reduce our stress levels. Interesting. (Has anyone investigated the effects of meditation while looking at fractals?)

There was a lot of money on the line in this particular situation, and apparently the politicking and public flame throwing were severe. As Taylor said in his article: "For the first time, computers were playing a significant role in determining the fate of artworks". The project was kept secret in order to protect its objectivity as well as Taylor himself. The full story reads a bit like a spy thriller. At one point he was advised not to go outdoors in case a bird dive bomb his head; at another point he was advised to he leave the country for a short while! Eventually his analysis, later supported by other evidence, determined the paintings were fakes. (This was about the time he was told to flee to New Zealand). The process of using software to analyse fractals in artwork is now legitimized and used by many researchers.

Reading this article caused me to think of a similar activity on a smaller scale and closer to home. My father (who passed away in 2009) was an archaeological chemist and spent hours on end sitting at his computer developing FORTRAN code to analyse art. Over the years, he was called upon to authenticate ancient ceramics, paintings from the Middle Ages, and was always glued to stamps (he'd appreciate the little joke). Although the culture of ceramics and artwork is interesting, it was his analysis of stamp forgeries that was most unique as far as I (in my admittedly biased position) could tell. He was always programming programming programming and peering at these little bits of lines and ink. I never got to see his code, and I suspect it is now lost, but I'd give a lot to know more about what he was doing. Over the years I would periodically be ordered off on a mission to obtain some obscure piece of software that he wanted to use, or to locate a specific piece of share-ware that he had somehow learned about. Fortunately he published more articles than I could ever read and a few books, so I can read his results even if I cannot see the code that produced it. If those subjects interest you too, here are a few books: on ceramics (here and here), stamps (here and here)

Sunday, December 12, 2010

Do you want to have a fight with your car's software?

Perhaps because it is Sunday and I had promised myself not to work today, yet find myself on the computer (again :) a little bit of irreverence and levity. Besides missing out on the sunny skies to plop myself in this ergonomically correct chair, I am working with 9 fingers, having accidentally smashed one of them into a steel pipe last weekend. Typing has been a bit challenging. But not nearly as much as flossing. Try flossing without both pinkies. It can be done but it takes creativity that would make my dentist proud.

I just listened to the podcast from National Public Radio's Science Friday. The host interviewed the high school pair who won the grand prize in the Siemens Foundation Math, Science and Technology contest  in.... yes... computer science. (The grand prize also went to someone working in astrophysics. NPR did not mention this, but it is on the Siemens Foundation contest web site. See above link) The prize is $100,000.  The host noted afterwards with some regret that he forgot to ask the teenagers what they would do with the money. I don't know...when you were 16, what would you have done with $100,000? That's a lot of [fill in your own blank].

The students developed code to identify emotions from your voice. The NPR host started off by describing how he would sometimes tell his automobile GPS to "shut up". And wondered if this technology would allow the car to adjust if it noted that he was pissed off. My first thought (remember, it is Sunday) was - so turn off the annoying GPS. It is bad enough when Google Maps takes me 5 miles out of my way and dumps me in the far end of a dead end parking lot when my desired destination was half a mile from where I originally got into my car in the first place. I have been in cars with people who cannot seem to drive without their GPS - brain, out the window. Annoying computer voice at the helm. ick....

So I'm not sure that I'd call that particular application of emotional recognition software socially beneficial. I might be more likely to get into an escalating argument with my car as it tried to be soothing and I became more and more annoyed at its cloying fakeness. Then I might find myself and my car in a ditch. Definitely not socially beneficial. Guess I should be clear that I'm pretty sure the car GPS application was the invention of the NPR host, not the students.

The students actually had something much more interesting in mind (assuming that your idea of fun and excitement does not involve fisticuffs with your dashboard). They are now considering creating a wristwatch that can be worn by Autistic children to help them in social situations. The watch will identify the emotion of the person being spoken to and show a picture (smiley face, frown, etc) to help the wearer correctly judge the mood of the other person. Now that is a bright idea and much more interesting I think. I hope they do it.

Meanwhile, I shall attempt to go back to not working....happy Sunday to all!

Thursday, December 9, 2010

"You Can Do Whatever You Want, But You Can't"

That is what a high school student told me yesterday. I was sitting in a room speaking with a few of the students attending Coleman Tech Charter High School, which I first reported on last August. At that time, the walls were just up, the cables had only recently disappeared into the ceiling, and the technology infused curriculum was new and shiny and ready to be tested. Students were being recruited and excitement was in the air. Nearing the end of their first term, I decided it was time to see how things were going. So I made a visit to Coleman Tech and spent half a day watching class, speaking to Vice Principal Neil McCurdy ("Dr. McCurdy" to the students) and several of the teachers. And of course the students.

The consistent message I heard from everyone was that the school was special; it provided a personalization uncommon for a public school and everyone had a story to tell to prove it.

The student who made the statement above was summing up several stories about why he likes the school. He says that the teachers and administrators (who overlap, as the qualified administrators also teach) really are open to new ideas and suggestions and pay attention to what students have to say. He feels like a person, not a kid afloat in a mass of bodies. That is the "You can do whatever you want" part. Another student concurred, stating that the school she would have attended has a class size of 60. 60??? Not sure I believed I had heard correctly, I asked for confirmation and she said, yes, really, 60. Eek. This student, who had zero background in computing, is blossoming at Coleman Tech, taking among other things, Computing 2, and doing "extra things" at home with her laptop.

The "But you can't" referred to the way in which discipline and enforcement of respectful behavior are handled. One story, told by another student, illustrates this point. This student felt bullied by a classmate. In that student's old school, said the student, the approach to dealing with the situation would have been for the teacher to make a public display of disciplining the student. Here at Coleman Tech, the student related, nothing happened that instant (note that the situation was not endangering in any way) but over the next few days the perpetrator began to act increasingly respectful. So, the student inferred, something had occurred outside of class. The long term effect was that the student relating the story felt more comfortable and a "scene" had been avoided.

Beyond agreeing with the student's assessment that someone had spoken to the misbehaving student outside of class, I also inferred that one of the reasons that this approach was successful over the long term was because the perpetrator had not been publicly humiliated, which can lead to increased behavioral problems and possible retaliation in other settings. The student who felt threatened was not forced to deal with whatever his peers thought of him - which in high school is a very big deal.

Coleman Tech Charter School has attracted students from all over the region. Some students come from "North County" which is outside of San Diego city proper, whereas other students come from inner city areas in the heart of San Diego. The students are a good representative of San Diego: multi-racial, economically diverse, and, close to 50-50 gender split (the girls slightly outnumber the boys in this tech high school!). I watched an incident unfold that demonstrated another way that Coleman Tech is unique and personal. Everyone, students and teachers, eat lunch together in one large room. At one point an altercation almost broke out and Neil McCurdy, as Vice Principal, stepped in. It turned out that one student came from a background where the response to a perceived provocation was to become overtly aggressive. The other student, from a different background, felt that increased provocation was appropriate. When Neil stepped in, it became an opportunity not only to enforce discipline, but to discuss why each student's approach was not going to get the student what they wanted. Nearby students also heard the discussion. Neil told me later that part of what they try to do at Coleman Tech is consciously teach students about cultural differences among each other and to learn how to successfully interact with people from different backgrounds.

Ok...where is the computing? Everywhere. Every student, as promised, has their own laptop and every class uses the interactive white boards. One teacher in a non math/science discipline at first told me that he didn't really use technology in his class, but moments later was describing how the final project for his class was to create a digital movie! In addition, every 9th grader (approximately age 14) must take Computing 1, which is a programming based class using, guess what, Alice (the same system used in the APCS Principles class that I have been following all fall). There is also Computing 2, 3, 4 and 5. Currently two sections of Computing 2 are running and students are focusing on 3D graphics and animation although when I sat in, one section was discussing how the internet works at the level of IP addresses, packets and routing. Students were fascinated to see live, via a "ping" that a message sent from their school to UCSD a few miles up the road, was routed through Los Angeles. It made for fascinating conversation.

The other section of Computing 2 was engaged in analysis and debate of the WikiLeaks controversy - one group was "the western nations" and the other was WikiLeaks; they each had to decide what technical measures they would take to both attack the opposition and defend themselves. Although the ideas started off with some (to be expected) adolescent suggestions such as bribing North Korea to nuke the United States and its allies, the conversation began to become more serious as the students realized the serious flaws in this kind of argument. Unfortunately class ended just as things started to get really interesting and I wasn't able to hear how it all turned out - or rather will turn out, as this will be a multi-day exercise. I noticed that even amid the joyful chaos of a group of 9th and 10th graders throwing out ideas at random, students were constantly experiencing the "lightbulb effect". I saw several students, girls and boys, saying things like "oh! could that actually happen to my program?" "does the internet really work that way?" "you mean that they can do THAT to my computer???". Minds were engaged.

Friday, December 3, 2010

Socially Meaningful Student Projects

Yesterday I watched some of the student final projects for the APCS Principles pilot class. The students had been given the option of submitting their projects into a contest in which their classmates would vote on the best one. Participation in the contest was completely optional but quite a few students took the plunge. Ten projects were selected for viewing in class and being voted on. They were pretty darned cool if I say so.

This final class assignment (which all students had to do) asked them to create a digital project to communicate their views on some issue facing society; they were given a wide ranging list of topics to choose from and if they thought they had a better topic they could propose it. One topic, acknowledging the large number of psychology students in the class, was to design an interactive psychology experiment.

The contest proceeded in round robin pairings, with students using their by now well finger-smudged clickers to vote. They were able to see the results of each round live. To those who know Alice well it was clear that significant programming sophistication, such as complex looping and list structures, went into some of the projects . A few students figured out how to use concepts that the class did not introduce, such as variables and button input boxes.

One of my favorite projects involved a group of penguins who emerged from an igloo to play and pick up trash. They spun, circled, danced, flapped their little flippers (do penguins technically have wings?), sometimes in a chorus line and sometimes going off to express their individual selves. One happily slid around on its tummy over the ice. Along the way they picked up and tossed trash towards the trash can. Too cute. And there were clearly lists, parameters and other well designed code structures underneath the fun.

Another project was more serious, without being too heavy, which came through as instructor Beth Simon had the honor of "playing it". The tree she was trying to save from destruction did not make it. There were all these objects (zombies, flying anvils) attacking the healthy tree while little bunny rabbits hopped around it. The player had to shoot the nasty objects to save the tree. Unfortunately, the tree withered (i.e. the player missed too many fast moving objects) and the forces of environmental destruction won out ... this time.

One project replicated a famous real psychology experiment in which a group of seminary students on the way to a lecture about ethics and helping people mostly ignored a passerby in distress (it was a "plant" fortunately). The student brought the scenario up to date, using the setting of a student late for class who witnesses a disaster and has to decide if it is worth risking being late for class and possibly hurting their grade by calling for an ambulance. Beth acted as the psychology subject, and asked the class what to do. They ended up running the scenario with both decisions (help and not help).

There were so many more fascinating projects. One used creative graphics to teach quadratic equations, another challenged the user to decide if they should respond to provocation aggressively or attempt reconciliation, another had a kangaroo trying to hop over a tree as it answered questions about tuberculosis, a snowman trying to eat only healthy food, a plea to help out at this holiday time by donating toys to children in poverty, and a fun question and answer animation about sharks facts (ever seen a shark spin around on itself?) and finally a darkly amusing story about a basement hacker who goes phishing and lures in an unsuspecting victim to perform identify theft. This story won the overall class vote, although the stage by stage competition was at times quite close.

Very very few students were texting, or using Facebook in class. This was fun and informational for everyone and a great way to approach the end of the term.

Wednesday, December 1, 2010

Alice -> Excel in the APCS Principles Pilot

Midterm follow up, but I'm afraid there isn't anything earth shatteringly new to add. As Beth Simon had promised, the (34) questions on the midterm looked very much like the in-class quiz questions. If you have been following the podcasts, then you know exactly what those questions look like. There were exam questions on virtually all the major topics covered in the course.

Having started to tally the types of topic questions, I began to decide that this effort probably wasn't going to help out those of you who are still interested in details about the excellent midterm results. I believe that the varied analyses of the project that will take place after the course concludes may produce more engaging information than if I list out a topic frequency count.

It is more interesting to briefly discuss what happened in the class sessions in which Beth discussed Excel. As mentioned in a previous post, she went to great lengths to make a smooth transition between Alice and Excel, showing the relationships between two seemingly disparate programs. Alice is an animation oriented programming system and Excel is a spreadsheet program, albeit a now quite sophisticated piece of software. However, in working through complex concepts such as relative vs. absolute addressing (often a tricky distinction for learners), Beth demonstrated through example how an Alice exercise (such as a singing group called The Beetles - no that is not a typo, they were insects) had underlying code similar to formulas they could create in Excel. She discussed similarities and differences in concepts such as loops and conditionals. All very creatively.

This type of teaching supports what the learning literature describes as Transfer - the ability of a learner to successfully apply understanding from one learning experience to another and to rapidly learn new related information. Observing the class it appeared that transfer was likely occurring because students asked intelligent and thoughtful questions about the Excel exercises and remained engaged throughout the lectures.

This is also a good time to note that the course has been supporting two other types of well researched learning that I have not specifically referred to previously:  problem based learning and collaborative learning. If you are not familiar with the research in these areas, an excellent read is How People Learn put out by the National Academies Press. It has become a classic and very accessible reference on learning in educational settings. Well worth the read.