Archive for November, 2007
Friday, November 30th, 2007Via John Langford:
A couple [of] security researchers claim to have cracked the Netflix dataset.
Apparently, they were able to link some anonymized users in the Netflix dataset to named users on the IMDb website. That’s a pretty interesting result. This is the first time I can remember seeing research on something other than collaborative filtering based on the Netflix dataset, and really I’m surprised it didn’t happen earlier. This is clearly going to be a really important resource, and I’m excited to see what else it turns out.
Marathon runner and ESPN.com Page 2 contributer Jeff Pearlman has what he claims is “the greatest 26-song running playlist of all time.”
Twelve weeks later, I present to you The List To End All Lists. It comes from hundreds of hours of research: digging through Internet sites; speaking with musicians and athletes; combing through songs ranging from Dio’s “Holy Diver” (solid) to Jeff Beck’s “Space for the Papa” (too long) to Hanson’s “Where’s the Love?” (surprisingly good).
I like some of the choices a lot (“Lose Yourself,” “All These Things that I’ve Done”), and some less so (“New Sensation”? I’m supposed to run to INXS?), but clearly we can do better. So I’d like to ask that everyone contribute his or her own top five running songs. Mine below the jump. (more…)
Of course I am, right? As noted elsewhere, I’m a terrible sucker for a Gaiman yarn. I also read the Heaney translation last year, and thoroughly enjoyed it. So, that’s what I’m pushing for when we make our annual post-Thanksgiving trip to see a big blockbuster. Ruth, however, does not have my implicit faith in Neil’s good sense, and so is skeptical of the whole “Angelina Jolie as Grendel’s hot mom” bit.
Luckily, the internets are full of thoughtful people to whom the whole thing made perfect sense, such as Nature editor Henry Gee:
What the film does is very clever: it assumes that the poem that has come down to us is a bowdlerized propaganda version (which it assuredly is, having been through several scribal hands since its original composition) – and proceeds to tell us what really happened. In so doing the script exploits all sorts of odd foibles in the text, showing that Gaiman and Avary well those passions read, stamped on those lifeless things.
[... T]he film establishes its integrity by following the story almost line for line (allowing for the usual compressions of adaptation) right up until the point at which Beowulf has to go looking for Grendel’s mother.
[...] She proves indestructible, and Beowulf can only escape by making a faustian pact in which she will grant him, in effect, eternal life.
When I saw the trailer I was inclined to dismiss this as bunk. But a closer reading of the poem revealed two crucial things.
You’ll have to visit Gee’s blog to find out what those two things are. But the upshot is that, beyond there being a good reason for the decisions made with the script, there is actually something interesting going on there.
(PS, I got both of those links from Neil’s blog, where he said they were the two reviews he’s seen which got the closest to representing what he and Roger Avary where trying to go with the script. Oh, and he also likes the Kindle, and dismisses the DRM issues out of hand, which frankly is just silly.)
Jorge Cham, the author of PhD Comics, spoke at UCI last year. Ruth and I were really excited, because it happened just days after we arrived, and (being big fans) we thought this was fortuitous. Unfortunately, we weren’t all that impressed by the talk itself. Dr. Cham was almost unrelentingly negative, without it really being clear what was exaggerated for humor and what was genuine enmity. Perhaps this was a shortcoming on our end, an inability to perceive the nuances, and a reluctance to countenance anything which would discourage us during our first week as graduate students.
But, since then, I read a lot of the comics and I wonder if Cham really “gets it,” where “it” is some vague quality of what makes the pursuit of science so important to people. Take, for instance, the most recent PhD comic, about visiting with astrophysicists. It’s the second half of a two-parter, the first of which got a favorable link from Cosmic Variance, and the people there know a lot more about this stuff than me. The “Beautiful all the way down” bit is nice, but check out this corner of the comic:
You don’t have to be an astrophysicist to understand that this kind of agreement between prediction and observation is just ridiculous. To ask “That’s it,” when “it” is so impressive, just seems silly. Contrast Cham’s response with a comic by someone who definitely “gets it.” Randall Munroe is responsible for the consistently fantastic xkcd. I could be wrong about this, buy my impression is that his comic really blew up around the time that people like P.Z. Myers linked to this comic:
This isn’t to say that PhD isn’t still pretty funny. Every lab on my floor has one or two of the comics tacked up on a wall. But I just don’t understand what Cham thinks he’s proving to whom with these (real or feigned) attitudes.