Archive for May, 2008

2008 Semantic Summer Internships

Wednesday, May 28th, 2008 · Kendall Clark

Our interns have been, as a group, quite remarkable: two of the five have completed interesting dissertations within a year of finishing their internships, and both of them got good jobs in industry (a cutting-edge startup and Oracle)—in fact, we tried to hire both of them, but apparently people who intern at Clark & Parsia are too special to work at Clark & Parsia. We’re the anti-google! :>

The other two are both pursuing PhDs, one in Manchester and the other in Prague. We’re expecting big things from both of them. Maybe the quality of the internship program will slip a bit and we’ll be able to hire these interns. Fat chance!

Markus Stocker, the fifth intern, who’s presently still working with us, continues to impress with hard work, intelligence, and good nature. He and Mike Smith are responsible for Owlgres, and now he’s added our XACML policy analysis suite to his oeuvre.

Like the man said: “and proud we are of all of them”.

Which brings me to this year’s Semantic Summer, our oh-so-cute name for our year-round internship program. We’ve put up a description of this year’s topics and areas of focus, which you can read on the Semantic Summer page.

If you’re interested, drop us a line.

At the Feet of the Master…

Monday, May 26th, 2008 · Kendall Clark

Enjoying the Kierkegaard Monument in Copenhagen

SemTech 2008: First Thoughts

Thursday, May 22nd, 2008 · Kendall Clark

Hard to believe, but we just finished our first SemTech conference. We were heads-down last year and didn’t want to attend till we could make a bit of splash, which I think we accomplished this week. We gave three talks—as I’ve said on the blog endlessly—and they were all very well received. The first two, on Pellet and Pronto, were rather surprising, since attendance was unexpectedly high (150 and 115, respectively). None of us were quite ready for that, least of all Pavel, who’s a bit shy, but will turn into a very good speaker with more practice.

Evren’s talk was something of a milestone for the SemTech community, IMO, since all the self-appointed “cool kids” have been hammering away for what seems like years about the Semantic Web—having 150 folks clamoring for information about OWL DL and Pellet suggests that the Semantic Web is not as irrelevant as the cool kids like to suggest.

Mike’s talk on XACML-DL wasn’t as well attended for various reasons (conference organizers smushed us with another good talk, but I don’t think the smushing helped us or Tim at all), but it went very smoothly (I don’t think any of us knew Mike was such a good speaker!), and we had real interest from folks in the audience.

Even more unexpected, I think, people were very interested in Owlsight and Owlgres, despite the fact that neither of them was really on the conference agenda. That’s a good sign.

Two quick thoughts:

  1. There’s no substitute for being there to make connections with people, both biz and personal
  2. We’re still in “early days” for this market (IMO, at least) and I found that encouraging since there is increasing growth and maturity (which is lumpy, to be sure, but still real)

Evren and I are off to Denmark for the next 10 days for a customer engagement; but I’m looking forward to following up with everyone we chatted with about collaborations, mutual support, etc.

Good stuff.

Personal Arguments

Tuesday, May 20th, 2008 · Bijan Parsia

Lest anyone think we’re all hard core technical stuff and geekery, here’s a nice touchy feely post!

I am, by training and inclination, a philosopher. A common feature of philosophers is that they like to investigate arguments, indeed, are passionate about arguments. That is, we like to present, explore, pick apart, and play with justificatory structures. Many of us can get quite worked up about them as well. So, with a philosopher like me you can have a passionate disagreement even when you agree on the conclusion!

(Yes, my beloved one is a hero.)

Of course, everyone knows that attacking people is, at best, bad argument form in most cases, not to mention impolite and counterproductive. Sometimes it is straight up necessary (that is, not all ad hominems are fallacious): For example, consider an argument that a certain person is habitually a bad actor. It’s impossible to separate the argument from the person involved since, after all, one is discussion that person.

However, one might think that in most technical situations that personal attacks are inappropriate. Of course, sometimes they are funny and lighthearted, sometimes they just slip out, and sometimes they communicate important information (like that an interlocutor is at the end of their rope). In the latter two cases, apologies are almost always appropriate, and in the first case, if people take it the wrong way, apologies are likewise appropriate.

I tend to believe that “attacking the idea” is ok. I know this isn’t generally true: With children, for example, it’s often not good. Also, people identify closely with their ideas. (I personally get upset if a cherished idea is shattered. It’s certainly possible to increase that upset with certain tones.) When they do that, they tend to read attacks on the idea as attacks on them. In that circumstance, a “dishing back what you got” response is pretty common (even if that’s not what you got).

One line that I tend to fall into is that I’m perfectly innocent since I merely attacked the idea, whereas other people are out of line because they attacked me. This is coherent, but, y’know, doesn’t work too well. It’s pretty infuriating to feel attacked, then to lash back, then to have the fact that you made a personal attack rubbed in your face. Even if you recognize that you escalated it is very natural to deny that you did. It’s also very natural to lash out harder. Indeed, people are very good at avoiding the recognition that they are doing wrong. After all, it sucks to be a schmuck. To be a mean person is worse.

I used to think that once I got to “the serious place” (college, grad school, faculty) that the “attack the idea, not the person” thing would become the default attitude and everything would be nice. Or vigorous, at least. But it sure doesn’t seem that way. I would prefer, for my own ease of behavior, that I could just walk through the world concentrating on having a thick skin with regard to my ideas. Alas, that really doesn’t suffice.

Furthermore, it’s not too hard to subvert the rule. It’s very easy to make a personal attack that is hidden as an attack on ideas. Or to make a personal attack without mentioning the person. (I had someone claim that they did not personally attack me because they did not mention my name, but merely pointed to three email that I had written as examples of alleged reprehensible behavior. Bleah!) It’s possible, then, to mistake someone’s attack on your idea as a disguised personal attack.

My main strategy for that is to make my personal attacks as clear as possible! Then I can argue that of course my idea attack wasn’t a disguised personal attack because if I’m going to attack you…you know it. This, again, is not the most winning of strategies, but has a somewhat appealing coherence. If you find coherence appealing, of course.

One place where I am especially thin-skinned is when people, however obliquely, call into question my honesty. (I’m not too fond of attacks on my competence either.) This doesn’t seem to bother some other people in the same way, especially if they can indite themselves in the same breath. For example, I had someone point to me after an exhausting thread (wherein I did my best to keep cool and argue the merits) and say, roughly, “Oh, you’re like me…you just enjoy the disputing.” I found this amazingly hostile. I was arguing in good faith and, by their own words, they were not. In the confession of that, they accused me of likewise arguing in bad faith. That’s so frustrating that I’m likely to go ballistic in response. I imagine that my doing so is surprising to many people, including the interlocutor, since, from their perspective, they are being friendly. I’m not sure why anyone thinks that their indifference to a moral flaw is likely to be shared by anyone else.

Although, I do try to admit to flaws or the possibility that I’m the problem (or my style is or…) in order to get people to calm down. That doesn’t work very well either a lot since a bad faith or sufficiently upset interlocutor will just take that as an additional stick to beat you with.

I find myself walking away from conversations a lot more, which makes me sad. I often come at things from a different perspective that people have trouble grasping. (A prof in grad school kept beating on my papers as heroically missing every point until I had a sit down with him and went through one bit by bit. He had a revelation: I wasn’t missing the point, I just was coming from a very different perspective. More precisely, “I saw the world askew.” This didn’t make me wrong, just harder for him to get without care.) One of these is the methodological and another is what constitutes appropriate, helpful, and polite behavior. I’m not very good at successful modification of my settled behavior (that is, I can effect fairly radical changes, but not in ways that help notably in the broad scheme of things) so it seems unlikely that I can unskew myself.

This leaves me with trying to compensate: cultivating an extra-thick-skin, getting better at recognizing when a situation is hopeless, continuing to practice de-escalation tactics, doing as much good as possible, and avoiding problem cases. Oh, and, of course, finding communities that I can live with and that can live with me.

3 Talks at SemTech This Week

Sunday, May 18th, 2008 · Kendall Clark

If you’re following us on Twitter (and if you’re not: for shame! :>), you’ll already know we’re giving 3 talks this week at SemTech, a big industry conference in San Jose. If you’re going to be there, you should try to catch all 3:

Which together offer a representative slice of what we’re working on these days: OWL DL reasoning, a core biz for us; policy management, which is up-and-coming; and probabilistic reasoning, somewhere between the two.

Looking forward to chatting with lots of folks this week. Safe travels, everyone.