Center for Machine Learning and Intelligent Systems
About  Citation Policy  Donate a Data Set  Contact


Repository Web            Google
View ALL Data Sets

KASANDR Data Set
Download: Data Folder, Data Set Description

Abstract: KASANDR is a novel, publicly available collection for recommendation systems that records the behavior of customers of the European leader in e-Commerce advertising, Kelkoo.

Data Set Characteristics:  

Multivariate

Number of Instances:

17764280

Area:

Life

Attribute Characteristics:

Integer

Number of Attributes:

2158859

Date Donated

2017-05-16

Associated Tasks:

Causal-Discovery

Missing Values?

N/A

Number of Web Hits:

11618


Source:

Massih-Reza Amini
Univ. Grenoble Alpes, CNRS/LIG
massih-reza.amini '@' univ-grenoble-alpes.fr

Charlotte Laclau
Univ. Grenoble Alpes, CNRS/LIG
charlotte.laclau '@' univ-grenoble-alpes.fr

Sumit Sidana
Univ. Grenoble Alpes, CNRS/LIG
sumit.sidana '@' imag.fr


Data Set Information:

We created this data by sampling and processing the www.kelkoo.com logs. The data records offers which were clicked (or shown) to the users of the www.kelkoo.com (and partners) in Germany as well as meta-information of these users and offers and the objective is to predict if a given user will click on a given offer.


Attribute Information:

userid offerid countrycode category merchant utcdate implicit-feedback


1. train_de.csv (3,14 GB)
Instances: 15,844,718
Attributes: 2,299,713
userid: Categorical, 291,485
offerid: Categorical, 2,158,859
countrycode: Categorical, 1 (de - Germany)
category: Integer, 271
merchant: Integer, 703
utcdate: Timestamp, 2016-06-01 02:00:17.0 to 2016-06-14 23:52:51.0
implicit feedback (click): Binary, 0 or 1

2. test_de.csv (381,3 MB)
Instances: 1,919,562
Attributes: 2,299,713
userid: Categorical, 278,293
offerid: Categorical, 380,803
countrycode: Categorical, 1
category: Integer, 267
merchant: Integer, 738
utcdate: Timestamp, 2016-06-14 23:52:51.0 to 2016-07-01 01:59:36.0
implicit feedback (click): Binary, 0 or 1


Relevant Papers:

Sumit Sidana, Charlotte Laclau, Massih-Reza Amini, Gilles Vandelle, and Andre Bois-Crettez. 'KASANDR: A Large-Scale Dataset with Implicit Feedback for Recommendation', SIGIR 2017.



Citation Request:

If you publish results based on this data set, please acknowledge its use, by referring to:
Sumit Sidana, Charlotte Laclau, Massih-Reza Amini, Gilles Vandelle, and Andre Bois-Crettez. 'KASANDR: A Large-Scale Dataset with Implicit Feedback for Recommendation', SIGIR 2017.


Supported By:

 In Collaboration With:

About  ||  Citation Policy  ||  Donation Policy  ||  Contact  ||  CML