Second Germanniversary: reminiscences

Posted in Uncategorized on 2013 September 1 by Asad Sayeed

It’s been two years since I moved to Germany, exactly, so I thought I’d put up a brief post in celebration of the event. I actually didn’t expect to be here this long, originally, based on the length of my original contract, although that was largely due to the peculiarities of academic funding that are the same everywhere.

I arrived early morning September 1, 2011, in Frankfurt airport. Had just a few minutes to figure out where the Regionalbahnhof was to catch my train to Saarbrücken, which I just barely did. (Wouldn’t have really hurt if I had missed it, but still.) First thing I did in SB after catching my bearings—I had an idea of the “top view” of the city from the Hauptbahnhof via Google Maps, so it was disconcerting to get the “side view”. So that’s what the Europa-Galerie really looks like!

I really had done a lot of research on Saarbrücken before even accepting the job, but some things have to be done with “boots on the ground”. I got a taxi driver to drive me to the temporary accommodation set up by the Max Planck Institute/MMCI for me, left my luggage there, and then somehow miraculously made it back to the city center to set up internet and cell phone service, which apparently I did miraculously quickly, according to some of my foreign colleages. The research as well as my basic proficiency with written German really paid off.

The next day I went to work for the first time, and was faced with the many days of on-going set up paperwork. Thanks to the good offices of the MMCI staff, I got my local residency registration straightened out (this isn’t immigration, even Germans have to do this—it’s peculiar to Germany as far as I can tell). Opened my bank account, started transferring money to it the old-fashioned way, through ATM withdrawals (I had favorable terms on this from my US bank—it’s not more expensive than using a transfer service, but there’s a withdrawal limit).

I even bought my first groceries from Aldi Süd—well, not really my first. Germans are often surprised when I tell them at Aldi Süd exists in the USA in mostly the same form as well as disguised as the more upscale Trader Joe’s. I recall that it was Quark (a form of German soft cheese not quite like cream cheese), bread, some fruit, and a frozen pasta dinner for which I also had to rush and buy some dishes to actually prepare. Also, milk—it’s only sold in max 1L cartons here, unlike the 1 gal. jugs that I was used to in the USA, or the 4L bags of Eastern Ontario that boggle even US minds. (You buy milk in thin plastic bags???)

Unfortunately, I was then plunged into a six week hectic apartment hunt in between actually figuring out my job and getting a start on research. This was partly my fault as I was very picky, but the tenancy system in Germany also has a lot of differences from the US and Canadian ones that aren’t altogether very favorable for very short-term renters. But you can’t have everything.

It wasn’t my first country change, but I still reminisce on the time. Even though it was only a few weeks, it was subjectively months long. In the end, despite all my research, I really only consider myself properly “set up” about three months later. That is defined as getting a home DSL connection, of course. Big German ISPs are no better than anywhere else, it was a Verizon-esque nightmare. Many people might consider getting their residency visa to be the endpoint of the adventure—that actually took a while because they were just introducing the electronic visa ID cards (that is an interesting issue…), but aside from the delay it was very smooth—being a Canadian has its advantages.

But while the whole set-up of my German life contained a lot of hassles, it was also a lot of fun. Like I said, I like to reminisce about it at parties. I can easily recommend a change of country as a way to shake up your life and wish it were in general easier for most people to consider it. I still haven’t moved to a country in which I am totally alien—as I said, I was German-familiar before moving to Germany, and all Romance-language countries are also more or less spoiled that way for me. A decade ago I hardly imagined I’d be living in the DC Metro area, let alone Europe, so who knows?

A year and a half later

Posted in Uncategorized with tags , on 2013 March 13 by Asad Sayeed

Been a while, eh? I’ve largely tended to post on Twitter, because it’s easier, and I can do it when I have an off-handed thought on the bus.  But German-language Tweeter and blogger 3c5x9 requested on Twitter a few months ago now that I actually write a blog update to my life in Germany.  Since it’s easy to procrastinate, thus I did.  I actually have a lot of thoughts, and if I actually don’t drop the blog for another few centuries, I might expand on them, but here are some highlights.

On the job

So I’m 18 months into this job, and I can say it was a very good choice.  If I may say so myself, Vera Demberg and I remain at the bleeding edge of computational psycholinguistics as well as at the intersection between theory, experiment, and application. We’ve defined for ourselves an ambitious problem: how to characterize the relationship between syntax, semantics, and overall cognition via the notion of “cognitive load,” the burden of cognitive processes on the language user.  Our main (but not exclusive) syntactic framework: Vera’s PLTAG formalism. Our semantic vehicle: Neo-Davidsonian event semantics.  Our experimental framework: eye-tracking, particularly pupil diameter as a measure of cognitive activity, and we use information-theoretic techniques to measure this from a language-internal side.  And our application: multitasking with language in an automotive setting.

We already have some interesting results on all these fronts, some published, some in the process of being published.  And on another note, I found I fit quite well into the intellectual culture here, give or take my in-Europe-unfashionable tendency towards Chomskyan minimalism.  I found myself adopted by the German comp. ling. community quite easily.  In a sense, some of my ways of thinking about things, particularly my interest in formal structure, never went as out of fashion as they have in North America—until recently.

Country

If you leave out the, uh, little matter of language, then Germany is basically 90% a Standard Developed Western Country.  People tend to work Monday-Friday, they draw cash from very familiar ATMs, they go to the movies or watch TV, some people eat at MacDonald’s, and soon, and so forth.  Salaried employees get a Lohnsteuerbescheinigung at the end of the year, which is just like a Canadian T4 or US W2 slip—good to know that tax returns, at least, are the same concept everywhere.

BUT, a lot of little details are contained in that 10%, especially if you’re coming from North America or possibly even English-speaking countries in general.  For example, the strange practice of Haftpflichtversicherung, widespread personal liability insurance.  It was only by chance that I even figured out it was necessary (prospective landlady trying to explain it to me and bugging out that I didn’t grasp the concept—turns out it wasn’t my German, I simply didn’t believe it).  The necessity of obtaining a Anmeldebestätigung from the Einwohnermeldeamt (residency registration from the resident’s office).  The widespread absolute Sunday closure of retail including basic supplies—unless you want to go to a gas or train station, where they will rip you off.  They have a special exemption because a carton of milk is obviously a necessary travel supply…

Not all differences, of course, are inconveniences, just so that you don’t catch me focusing on the negative.  During opening hours, the shopping concept is far cheaper and more convenient, at least for basic groceries.  The train network is heaven-sent, for a North American.  It’s a day trip for me to go to Frankfurt or Paris, and not terribly expensive either.  I have a nationwide smartphone plan and a USB data stick as well as unlimited DSL for a tiny fraction of what I was paying in the USA, for service that is no slower for my purposes and certainly at least as reliable. And the weather in Saarbrücken mostly suits me very well, very pleasant summers.

Some differences are simply neutral, and just have to be learned.  Now that I’ve had my current apartment for over a year, I’m learning the little differences in home maintenance.  You always say “Guten Appetit” before a meal with more than one person. (German eyeballs seem to fall out when I tell them that an equivalent custom doesn’t exist at least in North America.)  Little things like that.

Language

I had a bit of a head start on this, as I took four years of German classes when I was in high school in Ottawa.  This got me up to a fairly high level of “theoretical Deutsch”, meaning that I knew my Zustandspassiv from my Konjunktiv II.  It helps that I transitioned directly from there into linguistics, so that grammar has been at the forefront of my mind ever since.

Naturally, this wasn’t sufficient to become a participating member of society.  But it gave me a hook to progress quicker than circumstances would otherwise appear to afford—with help from the dearly-appreciated Max Planck Institute’s free language classes.  My current teacher has said that I could probably pass a B2 exam with a high score if I wanted to (language proficiency levels are standardized in Europe), and likely even C1—second highest official rank for foreign learners.  This does NOT mean that I am yet truly conversationally fluent, merely that in combination with some test-taking strategies, I can prove to the examiners that I am not an idiot, heh.

Conversational fluency requires actual conversation, and at work there are too many foreigners who don’t speak German for this to happen—what you get when you want to recruit highly-specialized researchers on the world market.  My reading proficiency is probably at C2 easily, otherwise, and I can make myself understood in written form.  But functioning in society really requires being forced to use the language, and very many foreigners who come here do not, especially when the workplace is international.  For me, part of the problem is confidence—I am a terrible pedant and am terrified of “hurting” the language through grammatical errors, like getting noun genders wrong!  Irrational, I know.

Still, there is a misconception that it is economical to live in Germany without German.  I strongly consider this a false economy.  Non-German learners simply forego many things, including things that could make their lives easier, cheaper, more convenient. Even a little bit of German proficiency can mean the avoidance of a stupid, expensive mistake. I may not yet be so great over the phone, but I am sure glad I do have some German experience.  (But, contrary to what you might think, my previous German experience was simply a happy coincidence—I didn’t choose Germany for German, but because it was a job I wanted.)

People and culture

Now we wade into thorny territory here, and what probably requires a separate blogpost.  So, let’s just say that brown people from English-speaking countries have a thoughtless belief that central Europe is white homeland of whiteness, where everyone is blonde and white—and many people I knew were sort of dubious about this aspect of my move.

Of course, for Germany at least, that hasn’t been true for a long time.  That someone is not exactly white doesn’t mean that their first language isn’t German or they weren’t born in a town outside of Stuttgart to German-speaking parents!  It is by no means all roses (or as Germans say, Friede, Freude, Eierkuchen), but…even if döner kebap comes from Turkey, it is as much a part of the fabric of German life as the stereotypical schnitzel, and what’s more, agonizing about what way it is part of German culture is also a part of German culture…

There’s room for a dozen blog posts here, and the truth is, German multiculturalism is more accidental and less planned/accepted/intentional than Canadian multiculturalism, and therefore not as far along.  But it is nevertheless a diverse society working on a ways to integrate its heritage into a demographically changing future.

And I’m a big fan of the subsidized opera and theatre, of German art and German museums, and the German temperament actually suits this Canadian quite well.

Food

German carnival food completely destroys American carnival food.  Reibekuchen (closely related to Jewish potato latkes), well, I’ll have them with or without Apfelmus. I wish Saarbrücken had more spicy food, but well, it’s a small city.

Politics

Once again, oh, boy.  Anyone who knows me well, knows that I am addicted to political and economic news.  Naturally, I am now addicted to German political news, and have Opinions.  (The capital O loses force for Germans, but anyway.)  That requires maybe a different post and maybe a pseudonymous blog.

But let’s just say that I’m a bit left of the political spectrum and am not a fan of German policy regarding the Euro and the Eurozone.  I do like German proportional representation much better than Canada’s Westminster-esque system, however.

Everything else and Zum Schluss

That only scratches the surface of what I haven’t (perhaps yet) blogged about.  As I said in a previous post, Saarbrücken is a nice, affordable city to live in.  It’s next to France, has great train connections in Europe, and friendly weather.  I made friends quickly here and got in touch with the anglophone expat community.  I could stand to find some activities in German other than MPI German classes, although time and, I think, a continuing lack of confidence have held me back.

Other themes I could have written more about: driving (people who prefer manual transmission are crazy, Germans prefer manual transmission, therefore…?), German television, shopping culture, the peculiarities of banking, rental contracts, and so on.  Maybe some other time.

Meine Forschung in Deutschland

Posted in Uncategorized with tags , , on 2011 October 7 by Asad Sayeed

In my previous post I talked about the act of moving to Germany, and how it has been so far.  But I suppose some of you will want to know about my work life at the University of the Saarland.

Well, I’m appointed in the department of Computational Linguistics and Phonetics (COLI), but in the way of academia, there are a series of bodies involved in my employment, and their relationships are something I don’t entirely understand yet.  I am apparently part of the Multi-modal Computing and Innovation Cluster of Excellence (MMCI), which is a joint project of a bunch of groups, including the Max Planck Institute and the Deutsche Forschungsinstitut für Künstliche Intelligenz (DFKI, the German Research Institute for Artificial Intelligence).  My supervisor is Vera Demberg, who is a Junior Research Group Leader appointed as part of the MMCI package.

I chose to accept Vera’s offer back in the summer and come here because Vera is very active in bridging the gap between the formal sort of linguistics which is near and dear to my heart (as some of you know), the psycholinguistics that I’ve always wanted to get my fingers into, and the more practical-minded statistical efforts to make systems that represent the world in a robust way, which is where I focused my PhD dissertation work.  So it was a close match of congruent interests.

Right now, we’ve been working on defining our actual research project and goals, which has actually been a lot of fun.  There are plusses and minuses to starting a postdoc when there wasn’t already an active project in place (Vera has other projects less technologically-oriented).   During my graduate career, I had multiple opportunities to define projects both for my thesis and for grant- and internship-driven work, and for the most part it worked out well, so on balance I’m pretty happy to be there at the beginning.  Another advantage is that I don’t have to reverse-engineer someone else’s peculiar code.  Well, for now, at least.

I’ve had some catching up to do on some recent developments in psycholinguistic and representational frameworks, and we’ve particularly been focusing on surprisal-based measures of cognitive load.  Surprisal in computational psycholinguisics has developed a burgeoning literature over the past several years, particularly stemming from foundational work by people like John Hale, now at Cornell.  As an information-theoretic measure, surprisal is one way to bridge the gap between formal representation and statistical robustness which other measures of cognitive load (and statistical modeling) do not do as well.  As long as we have a conditional probability in the denominator, we can find the surprisal at particular points in a string (assuming incremental parsing).  Then it becomes a matter of testing the predictions experimentally.

However, we’ve been looking at opportunities to apply these to various kinds of real-time information retrieval and user interface tasks, particularly with transcribed/ASR speech.  But, and here’s the catch, we want to augment the typically syntax-based surprisal measures with some kind of additional formal semantics, which will then allow domain-dependence in our applications.

We’re hoping not to build an entire experimental and processing pipeline from scratch, so we’ve been casting about for resources and collaborators, as well as looking for students from within the Saarland fold.  I suppose I’m biased, but the current state of psycholinguistics, statistical modeling, and syntactic/semantic formalism is coming finally toward an interesting convergence points where we can start to model natural natural language processing activities, so to speak, so I already see lots of exciting opportunities.  And I’ve only been here a month.

Das deutsches Leben

Posted in Uncategorized with tags , , , , on 2011 October 7 by Asad Sayeed

Guess it’s long past time for a blogular update.  I moved to Germany a little over a month ago, and a heck of a lot has happened since, and I have hardly had much time to think about producing more bloggy goodness, though I’ve kept my Twitter feed updated.  Spent the first week or so setting up my status as an immigrant to Germany and employee of the University of the Saarland.  Some of this is not even done yet; I have my appointment with the Ausländerbehörde (Foreigner’s Bureau) in a few days to get my medium-term residency permit (Aufenthaltserlaubnis) and foreigner’s ID.  I made the slightly foolish decision not to accept the housing offered by the university and look for something myself while staying in temp accomodations.   The search turned into a crash course in German real estate, which is quite different from renting in the US or Canada, let me tell you.  Even other Europeans apparently find it strange.

Living at about 50% German immersion has certainly forced my high-school German to improve.  Particularly my renting vocabulary.

Anyway, I finally found myself a place in the very heart of Saarbrücken with an excellent, unbeatable view.  When I get around to it, I might post pictures.  Ordered a ton of furniture; probably ill-advised, but I got tired of living completely like a student.  Even then, I mostly got the cheap stuff that will be easy to get rid of, since this is just a postdoc job.  So I have to go through much of the paperwork all over again to change my address with the city resident register—this is rather alien to people from English-speaking countries, the idea that the government has proof-positive of your whereabouts, and no, this is for all residents, not just foreigners—change my address with the bank, and so on.

So as for Saarbrücken itself, it offers a pretty high standard of living, and is not very expensive to live in.  Rents are less than half of what one finds in Washington DC.  It’s a city of about 200,000, but it serves a much larger metropolitan region, as well as being the main metropolitan centre for all of the Saarland, a good chunk of the Palatinate (the Pfalz in Rheinland-Pfalz), and a big chunk of Lorraine.  I have so far managed to make day trips to places in the Saarland and Lorraine; the farthest I’ve gone in Trier, which is well worth visiting.  I shall be back.

Now y’all are probably wondering what I’m doing at work, aside from filling out forms.  That, I will get to in the next post, if I have time to post it.  Unfortunately, with no permanent address, getting a stable Internet connection has been a challenge, and I’m blogging from work (and it’s pretty late here).

The current plan: goodbye Maryland, hello Saarland

Posted in Uncategorized on 2011 August 12 by Asad Sayeed

Hi folks.  Many of you know that I passed my dissertation defense.  Go me!  So now I am technically no longer a student.  It feels rather strange, because nearly every phase of my life has had one enormous overhanging Goal, whether it be graduating from high school, or my bachelor’s degree, or…  Now for the first time, I don’t have a single overhanging Grand Goal.  I can have multiple agendas, and one not necessarily dominating the others.  This will take some getting used to.  Well, unless I get a tenure-track job at some point.  In which case, I suppose the Grand Goal thing will come back, until I get tenure.

But I am getting ahead of myself, because you’ll now correctly wonder what I am going to be doing with my life.  Well, the first thing I am doing, once I have fully deconstructed my time in Maryland, is to move to Germany.  Yep, I am starting a post-doctoral position in Germany next month, at the University of the Saarland, working on incremental parsing formalisms with Dr. Vera Demberg, once all the paperwork is in order.  Change of continent, change of focus.

I’m more than a little sad to leave the DC area, of course.  The first three years I was here, I was but a tourist; if it had ended then, I’d have thought, “Huh, that was interesting.  The fourth and fifth years, the DC are became “familiar” to me.  These past couple of years, though, have left me with some amount of roots and personal attachment to the place.  It’s home, but the way of the world is that those of us with intended careers in research have to follow the jobs.  Even when times are good, I can’t just pick a city and expect to live there.  Doubly so for DC, where I am not a citizen and not eligible for most of the directly-attached government jobs. So I leave Maryland without the illusion that I would be coming back and settling down here.

But…a trip to Europe to pick up on a kind of work I used to do, and want to keep alive as part of my repertoire.  If I have to leave the land of the Smithsonian, there are few better ways I could have done it.  An opportunity to refresh my German too.

Now back to sorting through seven years worth of excess paper and packing and all those annoying details…

The grilling is scheduled a fortnight hence

Posted in Uncategorized with tags , , , , , , , , on 2011 July 20 by Asad Sayeed

Yep, my Special Day of Reckoning is approaching:

 THE DISSERTATION DEFENSE FOR THE DEGREE OF Ph.D. IN COMPUTER SCIENCE FOR

Asad Basheer Sayeed

Will be held:

DATE:                    Wednesday August 3, 2011 at 12:00 p.m.

LOCATION:           Room 3258 A.V. Williams Bldg.

TITLE:        A Distributional and Syntactic Approach to Fine-Grained Opinion Mining

ABSTRACT:  This thesis contributes to a larger social science research program of analyzing the diffusion of IT innovations. We show how to automatically discriminate portions of text dealing with opinions about innovations by finding {source, target, opinion} triples in text. In this context, we can discern a list of innovations as targets from the domain itself. We can then use this list as an anchor for finding the other two members of the triple at a “fine-grained” level—paragraph contexts or less.

We first demonstrate a vector space model for finding opinionated contexts in which the innovation targets are mentioned. We can find paragraph-level contexts by searching for an “expresses-an-opinion-about” relation between sources and targets using a supervised model with an SVM that uses features derived from a general-purpose subjectivity lexicon and a corpus indexing tool. We show that our algorithm correctly filters the domain relevant subset of subjectivity terms so that they are more highly valued.

We then turn to identifying the opinion. Typically, opinions in opinion mining are taken to be positive or negative. We discuss a crowdsourcing technique developed to create the seed data describing human perception of opinion bearing language needed for our supervised learning algorithm. Our user interface successfully limited the meta-subjectivity inherent in the task (“What is an opinion?”) while reliably retrieving relevant opinionated words using labour not expert in the domain.

Finally, we developed a new data structure and modeling technique for connecting targets with the correct within-sentence opinionated language. Syntactic relatedness tries (SRTs) contain all paths from a dependency graph of a sentence that connect a target expression to a candidate opinionated word. We use factor graphs to model how far a path through the SRT must be followed in order to connect the right targets to the right words. It turns out that we can correctly label significant portions of these tries with very rudimentary features such as part-of-speech tags and dependency labels with minimal processing. This technique uses the data from the crowdsourcing technique we developed as training data.

We conclude by placing our work in the context of a larger sentiment classification pipeline and by describing a model for learning from the data structures produced by our work. This work contributes to computational linguistics by proposing and verifying new data gathering techniques and applying recent developments in machine learning to inference over grammatical structures for highly subjective purposes. It applies a suffix tree-based data structure to model opinion in a specific domain by imposing a restriction on the order in which the data is stored in the structure.

Examining Committee:

COMMITTEE CHAIR:                          Dr. Amy Weinberg

Dean’s Representative:                      Dr. William Idsardi

Committee Members:

Dr. Jordan Boyd-Graber

Dr. Hal Daume III

Dr. Donald Perlis

EVERYONE IS INVITED TO ATTEND THE PRESENTATION PORTION OF THIS DEFENSE

Putting my stuff up where someone can see it

Posted in Uncategorized with tags , , , , , , on 2011 March 17 by Asad Sayeed

I just posted a draft article that I wrote with my advisor Amy Weinberg on LingBuzz, which has become a sort of joint working papers archive or respectable self-publishing system for generative grammar folk.  It’s a great idea—linguistics journals are hyper-competitive and who knows how many interesting ideas languish in the queue?

This article is a draft of something that I’ve been working on for several years.  It’s a kind of mini-linguistics-thesis inside my computer science degree.   I had to stop significant activity on it in March 2009 in order to devote the bulk of my time to experiments for my dissertation and other projects on which I was working at the time, so it was left in a (in my opinion) nearly ready state, but just needs a couple of more solid months of work, or a couple of years of hacking at it in my non-existent spare time.  But, you know, dissertation…

So I was faced with a choice of hiding my light under a bushel (or whatever the saying is) while I waited for that solid time-period or letting the world see it before it became a little too out-of-date.  I choose the latter.

What it’s about: The abstract (and article) is written in syntax-ese, but it’s the culmination of some thoughts I’ve had about languages that allow for long-distance extraction of noun phrase constituents from finite clauses…there I go with the syntax-ese again.  But let me put it this way: in some languages you can turn:

I now see that the jelly donuts were tasty.

into something like:

(Not English) The jelly donuts I now see that _____ were tasty.

But English does something weird as well: preposition stranding in questions:

What is he going on about ____?

In our article, Amy and I suggest that these facts are related, but current versions of generative syntax have removed the mechanism to express relationships of this kind.  These were called “escape hatches” in the past, and we propose a way to express them that is consistent with the terms of current theory.

I suspect/hope that it will be controversial.

Update 11:20 – 6 downloads and it’s hardly been 20 minutes.  That was fast.

Follow

Get every new post delivered to your Inbox.