Past Meetup

PWL#18 => Bob Poekert on COOLCAT

This Meetup is past

148 people went

Location image of event venue

Details

Mini

David Brockman (http://web.stanford.edu/~dbroock/Stanford/David_Broockman.html)on "What uncovering a massive academic fraud taught me about how academia needs to change". A bit of background

http://nymag.com/scienceofus/2015/05/how-a-grad-student-uncovered-a-huge-fraud.html

Here's what David will be presenting: https://drive.google.com/file/d/0B_Qj0otlErJqVlJtMUhTU3ZiRzQ

David's Bio:

David Broockman (http://web.stanford.edu/~dbroock/Stanford/David_Broockman.html) is an Assistant Professor of Political Economy at the Stanford Graduate School of Business. His research uses randomized controlled experiments to help understand public opinion, voter behavior, and political representation. His work on these topics has appeared in the American Journal of Political Science, Political Analysis, and other academic journals, and been discussed in the New York Times, Washington Post, and many other media outlets. David has also helped design hundreds of experiments for the field, the web, email, mobile, and mail in the academic, political, and private sectors, including with the Analyst Institute, CREDO, and Google.

Main Talk

Bob Poekert (https://twitter.com/bobpoekert) will present COOLCAT (https://cs.gmu.edu/~dbarbara/COOLCAT/coolcat.pdf (http://wing.comp.nus.edu.sg/downloads/keyphraseCorpus/58/58.pdf)) by Daniel Barbara, Julia Couto, and Yi Li.

From Bob "The paper presents an entity resolution algorithm based on minimizing entropy within clusters. Entity resolution (aka identity resolution aka categorical clustering) is a problem that comes up a lot in practice but which most people try to solve just by heuristics because they don't know that it's something that there's research on. This approach is pretty general (it applies to anything that you can represent as discrete field-value pairs), and the math isn't too overwhelming. Also I like the name."

Bob's Bio:

Bob (@bobpoekert (http://twitter.com/bobpoekert)) works at the intersection of data mining and databases at Factual, and got hooked on CS papers after reading the Kademlia paper in high school while trying to figure out how Kazaa worked. Bob also does weird stuff with image processing (@a_quilt_bot probably being the most popular example).

Meeting mechanics

Doors open at 6:30 pm; the presentation will begin at 7:00 pm; and, yes, there will be food.

After the paper is presented, we will open up the floor for discussion and questions.