Paper top 40 predictions

I've decided I am absolutely terrible at predicting which papers will be "a hit" and which papers will never get cited or read.

Like most academics, I'm usually working on several papers at the same time. In my mind as we're preparing and submitting, I often place bets on how the reviews will come back. Formulating my prediction involves not only the content of the paper, but also the publication venue, who I expect the reviewers might be, and other misc. variables.

I'm nearly always wrong.

And post-publication, the papers I am most proud of are never cited. The papers I am most ashamed of are frequently cited. I am considering employing reverse psychology as I write.

Either that or switch into mobile computing. Those people are so connected every paper has a gazillion citations.

Signal Boost: Stop the Internet Blacklist Bills

If you went to google.com, boingboing.net, wikipedia, or dozens of other sites on the net today, you may have noticed they have been blacked out in protest. This was done to bring the public's attention to two bills before congress: SOPA (Stop Online Piracy Act) and PIPA (Protect IP Act).

Why the protest? Well, these bills were intended to curb online piracy and copyright infringement (good), but did so in a really technologically uninformed and dangerous way (bad).

In addition to these bills not actually helping curb online piracy, they grant an incredible amount of leeway to allow the government and companies to arbitrarily censor and monitor the communication of people using the internet, both in our country and abroad. A few fun nuggets about the PIPA bill, quoted from publicknowledge.org :

• PIPA is overbroad. By including "information location tools," it makes nearly every actor on the Internet a potential violator.
• PIPA is bad international precedent. By sanctioning government interference with DNS, it would be used as justification for other countries to hinder freedom of expression of online.
• PIPA is ripe for abuse. By creating a "private right of action," rights holders could directly go after payment processors and ad networks.
• PIPA speeds fragmentation of the Internet. By targeting DNS, it could lead to a fragmentation of the Internet, running contrary to the U.S. government's commitment to advancing a single, global Internet.

There is a lengthy list of reputable organizations protesting these bills, including legal scholars, human rights organizations, industry groups, and engineers. Also, Joi Ito, director of the MIT Media Lab and fellow CS blogger, has a great post summarizing this issue, as does Trevor Trimm of the Electronic Frontier Foundation.

I urge you to take action and urge your congress members to reject this bill. 

More thoughts on unsolicited professorial advice

One other thing you get a lot of advice about as a new professor is how to run your show.

"Don't spend too much time on teaching"

"Write every day"

"Don't take too many grad students your first year"

The thing is, like anything, do what works for you. You want to spend 14 hours on making Teh Perfect Slides for your first class, do it. You want to get up at 4am and start writing, go for it. Want to relax all weekend playing Facebook games while occasionally picking at your grant proposal, sounds grand!

The trick is to know what makes you happy and know your own style, and work that way. You have a lot of flexibility in your schedule, the trick is figuring out how to best structure it so you're most productive. And to factor in recharging time, for whatever that means for you.

Fun with Google

You know that game where you type something into Google and note its wacky suggestions? I have no idea why, but I tried it today with 'Professors are'. This is the result:

So it would appear I am a candy bar, overpaid, and super-mean. Excellent! Ready for the Spring semester to start.

I also tried 'Computer Scientists are', and this is the result:

I suppose 3/4 ain't bad (I kind of like the blank one...)

Merry Hackmas (not really)

While most technologists were busy yesterday drinking eggnog and trying out their new gadgets, others were busy hacking Stratfor, an intelligence news organization.  All the news outlets reported it was 'Anonymous', but now people are saying it was (apparently) Sabu from LulzSec.

(Frankly, I can't blame the news outlets for the error - I can't keep up with the drama of who's who any more. It's like a soap opera, really.)

Anyway, whomever it was, they hack into Stratfor, steal a bunch of credit card numbers of people who subscribe to the company's intelligence briefings, then a) post them on the internet and b) use the credit cards to make donations to charitable organizations.

I'm not really sure what the point of this is. Any of these donations will be returned, and all the credit card numbers will be canceled. Really this will just cost the credit card companies lots of money, which will just result in the average Joe/Joann having to pay higher fees. Exactly what people need in this economy.

I wish these hackers would do something useful with their time. Solve some problems on challenge.gov. Teach math and computer science to children. Help local governments have more up to date computer systems in order to help empower communities.

Anything, really. This is just a sad waste of tech brains.

Thank you, thank you

One of the best things you can do as a student or employee is write someone a thank you note after they've done something helpful for you.

I was pleasantly surprised to find a hand written note from a student thanking me for a reference letter I wrote for them. This was immediately followed by a thoughtful note from an editor thanking me for a review.

These two things made my day.

Kudos, student and editor!

NRC Computer Science Rankings Reprise

There's an article in CACM this month by Computer Scientists Andrew Bernat and Eric Grimson, Doctoral Program Rankings for U.S. Computing Programs: The National Research Council Strikes Out. It talks about the ways in which NRC rankings are broken for CS (we have heard this before), but it details ways in which it could be fixed, which we hadn't heard before, and I like.

Two suggestions I thought were good:

• "Explore making the rankings subdiscipline-dependent. It is clear that different departments have different strengths. Thus, enabling a finer-grained assessment would allow a department with strength in a sub-field, but perhaps not the same across-the-board strength, to gain appropriate visibility. This may be particularly valuable for students deciding where to apply."

• "Use data mining to generate scholarly productivity data to replace commercially collected citation data that is incomplete and expensive."

The first is a nice idea; for example, you might be interested in a top ranked department, but it turns out 19/20 faculty focus on Theory and you actually want to do Systems. Or there might be some school with three top faculty exactly in your subspecialty, but you don't see them because they're 93rd in the rankings. 

The second is nice as well; I think with Google Scholar Citations data available this turns out to be a trivially easy problem to solve. 

Maybe CRA can do their own rankings; they collect a lot of their own data anyway, and it avoids needing to rely on the NRC.