Profile cover photo
Profile photo
Chenhao Tan
Chenhao's posts

Post has shared content
I was unable to finish a recent New Yorker article (by John Colapinto) because it began like this: professor Andre Geim wants to study very thin layers of carbon.  A new PhD student, Da Jiang, tries his best, using up all of the initial crystal in the process, but Geim isn't satisfied.  "As Geim teasingly admonished him ... one of his senior fellows glanced at a ball of used Scotch tape in the wastebasket, its sticky side covered with ... graphite residue".  Geim realizes that you can use tape to pull layers apart, and "enlisted the help of a Ph.D. student named Konstantin Novolselov" and the two of them (eventually) publish in Science and jointly get the Nobel Prize. 

What was the name of the person who noticed the Scotch tape? 

It could be one of the six other  authors of the Science paper,  S. V. Morozov, D. Jiang, Y. Zhang,  S. V. Dubonos,  I. V. Girgorieva, or A. A. Firsov  (except, I guess, D. Jiang, given the way the story is told); if so, the noticer of the tape certainly got their due.  But if not, wouldn't you feel a little bad for them?

So I started poking around to see if I could find out who it was, searching for articles with direct quotes from either Geim or Nevoselov.  In a 2008 Science Watch interview, the story is that Jiang challenged Geim to do better, and "This was a point of no return. I decided to use Scotch tape".   A 2013 Independent says, "Novoselov says, they had almost given up with graphite when they heard about how microscopy researchers working along the university corridor used Scotch tape to clean the mineral before putting it under the lens. “It was not a new technique, and I’d heard of it before, but when you see it in front of you it makes it obvious what it can be used for,” Novoselov recalls."  USA Today said in 2010 that "Novoselov says the pair actually found their first graphene sheets on strips of tape used to clean off carbon block samples in their lab, in preparation for a more complex technique for creating graphene. The researchers would, he says, "pull the tape off, and then throw it away. So once, I just picked it up out of the trash, and we analyzed it."" E&T magazine  in 2011 starts with the bit about a student having polished a piece into dust; "Then technician Igor Sklerevsky demonstrated how to clean graphite. It was known as the Scotch Tape Technique...Geim then found a piece of oxidised silicon with oxide that happened to be just the right thickness".

And here I was about to conclude, "Ah ha!  The original story has been morphed so badly in the telling that by the time the New Yorker got to it, in 2014, an imaginary senior fellow had been added in to the tale!", when I found Geim's 2010 Nobel speech. Under the section "Legend of Scotch Tape" [the word "legend" seems significant]: "Oleg Shklyarevskii, a senior fellow from Kharkov, Ukraine was working nearby... He [brought] over a piece of cellotape with graphite flakes attached to it. Allegedly, he just fished out the tape from a litter bin. I looked in the microscope at the remnants of graphite (Figure 2) and found pieces much thinner than Da’s speck.  ... things started to look promising and required more people to get involved. Oleg did not volunteer to take on yet another project but Kostya did." [OS did score an acknowledgment in the Science paper for "discussions and interest".]

Friends, what is the moral of this tale? There seem to be several possibilities.  Is this just yet another example of the Rashomon effect, or in how stories are changed in the re-telling, by the press or by the participants? That your Nobel Prize might be in your trashcan right now, if only you would look into it in the right light?  That you should be very careful about turning down projects --- or very lucky about accepting the right ones? 

The possibility that it is just "I should spend less time posting about New Yorker articles on social media and get back to work" has crossed my mind.


Post has shared content
Thank +Ed Chi +David Huffaker +Gueorgi Kossinets +Alex Smola for the great summer.

The slides are also available now on this page,

We show that it is possible predict expert ratings from user ratings. Also, users tend to get harsher as they get more experience.
Our CIKM paper presentation was jammed packed today with people sitting on floors. The topic was how to improve expert and user ratings of restaurants using joint optimization techniques.

+Chenhao Tan did a a great job presenting the work, done with my co-authors +Gueorgi Kossinets +David Huffaker and +Alex Smola .

Via +Xavier Amatriain Tweet:

Post has shared content
We had a very nice discussion on interpretability in machine learning. Karthik offered great thoughts in this blog. I also uploaded the slides that I used.

Post has attachment
I mostly agreed!

Post has attachment

Post has attachment

Post has shared content
News from my Cornell "kids" this summer: Oren Kurland 'PhD 06 was tenured at the Technion, +Bo Pang PhD '06 moved to Google, +Eunsol Choi BA '12 is going to UWashington for grad school, and  +Cristian Danescu-Niculescu-Mizil PhD '12 will be starting a tenure-track position at the Max Planck Institute SWS after a one-year deferral to do a postdoc at Stanford.  And although +Justin Cheng BS '12 can't  be said to be my academic descendant, since we're coauthors I'll go ahead and mention that he's starting grad school at Stanford.  Congrats to all!!

 [Image credit:]

Post has shared content
ACL folks: just wanted to share this great image +Cristian Danescu-Niculescu-Mizil made associated with our upcoming ACL poster. Looking forward to seeing folks in Jeju!

Coauthors: +Justin Cheng +Jon Kleinberg 

Post has attachment

Post has shared content
Wait while more posts are being loaded