Dear guy: I can't extract the downloaded sample dataset. everyone meet this problem? Thanks in advance David...
30
Markus Weimer
markus@...
Mar 1, 2011 6:18 pm
Hello, there has been a small hiccup with the datasets on the website: The file for track1 carries a .bz2 extension, while it in fact is gzip compressed. We...
Hey Guys, I am not able to find the registration link for the KDD Cup. Can you please provide me the link or inform me where I can register myself? Waiting for...
32
Markus Weimer
markus@...
Mar 1, 2011 11:03 pm
Hello, the registration page should be open by now under: http://kddcup.yahoo.com/registration.php Take care, Markus...
To identify whether an ItemId is a song/track, album, artist or genre: Songs =(Tracks) are numbered 0-624960 Albums are numbered 9-624943 and "None" Artists...
Track 2: to identify whether an ItemId is a song/track, album, artist or genre: Songs =(Tracks) are numbered 1-296110 Albums are numbered 0-296109 and "None" ...
In the Track 2 instructions, there are 6 ratings for each User:: 3 observed high ratings 3 imputed ratings, assigned sample average values The challenge is...
36
Zhaoquan Yuan
zqyuanustc@...
Mar 2, 2011 4:36 am
-- Best regards, Zhaoquan Yuan ... National Laboratory of Pattern Recognition, Institute of Automation, Chinese Academy of Science, Tel: +86-187-0151-6100...
Dears, I am sorry I just jump in. Proly repeating a thread. But would it be possible to have details on the evaluation? Beside some exadecimal/real numbers...
39
Markus Weimer
markus@...
Mar 3, 2011 6:57 pm
Hi, what details about the evaluation are you after? We use RMSE for Track1 and the error rate for Track2. Take care, Markus...
Hi, one question, how can one be sure that the registration was correct ? The registration page displays some textboxes and a "save" button. When i return,...
42
Markus Weimer
markus@...
Mar 4, 2011 11:00 pm
Hi, thanks for noticing! We are investigating the issue. Take care, Markus...
43
Markus Weimer
markus@...
Mar 4, 2011 11:06 pm
Hi, if you go to the page and it is already filled out, you are successfully registered. You can confirm that by logging out of Yahoo! and then visiting the...
Hello, there is a description of the format on the KDD Cup website: http://kddcup.yahoo.com/datasets.php I believe the information you are after can be found...
Hello, In the FAQs section, it says: "We will provide a simple sample program that reads the data, produces a simple model and writes a file with predictions...
49
Markus Weimer
markus@...
Mar 7, 2011 7:16 pm
Hi, yes the sample code will be available at the time the competition starts, if not earlier. Take care, Markus...
We are looking forward to get our hands on this huge data set! When we had a first brainstorming session we came up with the question on how users selected the...
52
aragorn459
aragorn459@...
Mar 8, 2011 10:50 pm
Hello, A quick question on how the ratings were collected: It seems that some users have rated dozens of items on the same day, at the same time. For instance,...
Hi there How can a user rate about 200 different artists in the same minute (user # 1)? Is there more detailed information about the data collecting process? ...
58
epfl_ch
epfl_ch@...
Mar 11, 2011 2:40 pm
Hi, I think there is already an pending question on this: http://tech.groups.yahoo.com/group/kddcup2011/message/52 Best...