Monday, December 04, 2006

Slides from my talk at Stanford

I had the great pleasure of giving a talk today on practical issues in personalization and recommendations for the Data Mining (CS345) class at Stanford taught by Anand Rajaraman and Jeff Ullman.

The slides from my talk are available in two versions. The first version is the talk I actually gave; make sure to read the notes pages for the slides, or it will be difficult to follow. The second version is done in a very different style and should be easier to follow without me blabbing away in front of you.

It was very fun giving this talk. The students were clever, thoughtful, and enthusiastic. I was pleased to get a chance to talk again with Anand, a sharp former colleague from And, I was overjoyed to meet Jeff Ullman who, for many of us computer science geeks, is a legend because of his seminal work and books.

I hope the talk was as fun for those in the audience as it was from the podium. Thanks again, Anand, for inviting me to speak.

See also my earlier post about the excellent lecture notes that are publicly available for this data mining class. If you are working (or just dabbling) in this field, they are well worth your time to review.

Update: Matt Wyndowe, who was sitting in the class, posted some thoughts on my talk. Thanks, Matt!


burtonator said...

Oh man.... I would have liked to have been there. You should have posted it on your blog :-/

No podcast?

Anonymous said...

Very nice to know about your Stanford Lecture.

What would you suggest students (of both at undergrad and grad levels) , on what kind of problems they should be looking at in datamining?

Greg Linden said...

I am a big fan of recommendations on websites (e.g. recommending books, online vidoes, postcasts, etc. based on past behavior). Both useful and lucrative!

If you want to work on that as a student, you cannot ask for a better problem and data set than the Netflix Prize!

Anonymous said...

Your very welcome, Greg. Great talk! Hope to see you around Stanford again.

M said...

Hi Greg,
I enjoyed your talk, it was pretty informative.
And thanks for the additional insight you gave in your comment on my article :-)

Anonymous said...

Thanks ppts where really helpful. I liked the slide on page loading speed and sales comparison for Google and Amazon. Recently a similar survey was done for Online retailers.

According to the survey site pages should load in four seconds or less.

Anonymous said...

Found my way here reading Designing Data Intensive Applications (author cited this post). The link to your slides are broken. Can you reupload please? Thanks

Greg Linden said...

Hi, Omar. Sadly, Google in recent years lost interest in preserving access to information for all time on the web and websites like Google Sites and Google+ are completely gone, taking old links across the internet down with them. So much for "organize the world's information and make it universally accessible and useful." But that's why this 2006 post has broken links. It's a real shame.

If this old talk is of interest, I did find I still have the file, and I uploaded it again to another Google product that hopefully they'll keep around this time. Here it is! I hope that it is still useful to you.