jrtom: (Default)

found/discussed here: http://it.slashdot.org/article.pl?sid=06/10/17/002251

Not a bad overview article on the phenomenon. (Although the /. discussion seems to suggest that the people that annotated the screenshots didn’t really know that much about IRC; apparently the “lots of cryptic data” is fairly standard stuff.)

As the (provocative) title suggests, the outlook of those interviewed is not exactly rosy. But there may be other ways of dealing with the problem than the "whack-a-mole" method of trying to root them out individually.
jrtom: (safe cat)

I so don't have time for this right now, but it sounds like a really interesting problem. Not sure I'm eligible to participate (my PhD advisor is apparently one of the judges, which I think explains a conversation that I had with him a couple of weeks ago involving a lot of intentional vagueness on his part), anyway, but I bet that I could do a pretty decent job if I had the time to sink into it. (Especially considering the data that they have but are not using...)

jrtom: (Default)

Plus some other stuff about working at Google. Some insightful comments in response, too.
jrtom: (execute)
A recent "oops" in an announcement of a lecture to be held at work (the announcer accidentally used the wrong title) gave me the idea for organizing a lecture series for 1 April: invite N people, get the topics for each, and then the day before (or perhaps a week, depending on the effect you want), ask lecturer ‘k’ to give a talk on the topic proposed by lecturer ‘k+1 (mod N)’.

Clearly, choosing one's lecturers carefully (for sense of humor, a generally relaxed attitude, and perhaps a fondness for perverse challenges) is paramount here. But in all seriousness, with the right speakers you could get some fascinating (and hilarious) lectures.

([livejournal.com profile] danyelf_moop tells me that he used to be in the perverse habit of putting up radically wrong title screens for his talks ("3D rendering of natural light in closets", "Photorealistic ray-tracing on the Palm Pilot" [which was a lot funnier for a 160x160 monochrome display--you probably _can_ do photorealistic ray tracing on today's PDAs]) until the talk actually started. Could be a fun alternative until I get enough pull to be able to schedule an entire lecture series. ;) )
jrtom: (Default)
Women In Science

The title is actually somewhat misleading, as it has more to do with reasons why people go into science as a career. I think some of it is oversimplified, but it's worth reading.
jrtom: (Default)
One of the ironies of the data sets that I study--social networks--is that they are both omnipresent and often difficult to get access to. So my research has been driven, in part at least, by the properties of the data to which I've been able to secure access. Sometimes organizations (companies, e.g.) will make data sets available to those in academia, and I've benefited from this, but it doesn't happen often. (The fact that a bunch of Enron's corporate emails got dumped on the web has, no joke, changed the course of the field of social network analysis.)

It's just occurred to me that if I take a job in industry, this problem will, in a weird way, invert itself: the company that I work for may be able to give me all sorts of data to work with...but no one else will.

jrtom: (Default)
I just got an email from the organizers of an academic/professional conference in the field of social network analysis. The gist is that they have too many abstracts for the amount of time that they have, and are therefore trying to figure out what to drop.

In this email, we find the following gem:

It is likely that we will drop some papers from the program because they aren't about social networks, they don't make sense, they have obviously been lifted from the internet, or for some other reason that convinced us that they don't belong on the program.

Now, I realize that this conference has never claimed to have a formal peer review process for inclusion in the program; it's a conference to which one can bring work in progress, and generally work of a speculative nature. I'm generally fine with that; there's a place for such conferences, and I'm glad this one exists. Heck, I presented there last year and probably would be doing so this year if I had more time.

But I mean, seriously, have the organizers not at least been doing the minimal checking required to filter out papers that aren't about social networks, or don't make sense? (I guess this would be the "hemorrhaging edge" . . . . )


jrtom: (Default)

May 2011

1516 1718192021
29 3031    


RSS Atom

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated 20 July 2017 16:31
Powered by Dreamwidth Studios