yoochoose — YooChoose e-commerce sessions (RecSys Challenge 2015)

Description

A bipartite network of the browsing sessions of the customers of a European online retailer and the items they clicked on and bought, released by YooChoose GmbH for the RecSys Challenge 2015 (whose task was to predict which sessions end with a purchase, and which items are bought). The data were collected during six months of 2014 (April to September) by tracking the website of the retailer, whose identity is not disclosed; all numbers (identifiers, prices) were modified by the source to protect the privacy of the users and the retailer.

A directed edge from a session to an item is a click on the item during the session, or a purchase of it (edge property buy). The network is a multigraph, since items are often clicked several times in a session. Sessions are not linked to persistent users. It combines the three files of the challenge: the clicks of the training sessions (yoochoose-clicks.dat, 33.0 million clicks), the purchases of these sessions (yoochoose-buys.dat, 1.15 million purchases; most training sessions have none), and the clicks of the test sessions (yoochoose-test.dat, 8.3 million clicks, marked by the edge property test), whose purchases are not given.

The network has the following properties:

(Node) is_session: whether the node is a session (otherwise it is an item).
(Node) id: the identifier of the session or item.
(Edge) time: the time of the click or purchase, as Unix time (with milliseconds).
(Edge) buy: whether the edge is a purchase (otherwise it is a click).
(Edge) test: whether the session is one of the test sessions of the challenge.
(Edge) category (clicks): the context of the click: 0 if missing, 1 to 12 for one of the categories of the shop, or the 8-10 digit identifier of a brand (-1 for special offers, and for purchases).
(Edge) special_offer (clicks): whether the item was clicked in the context of a special offer or promotion.
(Edge) price (purchases): the price of the item (modified by the source; -1 for clicks; 0 for 610,030 of the 1,150,753 purchases, whose price and quantity are not given).
(Edge) quantity (purchases): the number of items bought (-1 for clicks; 0 where not given).

Tags

Economic Commerce Bipartite Multigraph Timestamps Metadata

Added

2026-10-28

Citation

D. Ben-Shimon, A. Tsikinovsky, M. Friedmann, B. Shapira, L. Rokach, J. Hoerle, "RecSys Challenge 2015 and the YOOCHOOSE Dataset." Proceedings of the 9th ACM Conference on Recommender Systems, 357-358 (2015) 10.1145/2792838.2798723 Sci-Hub

Upstream URL OK

https://recsys.acm.org/recsys15/challenge/
https://s3-eu-west-1.amazonaws.com/yc-rdata/yoochoose-data.7z

Upstream license

CC BY-NC-ND 4.0, https://creativecommons.org/licenses/by-nc-nd/4.0/

Networks

Tip: hover over a column header for a description.

Column legend

$\left<k\right>$: Mean degree
$\sigma_k$: Standard deviation of the degrees
$\lambda_h$: Largest eigenvalue of the Hashimoto (non-backtracking) matrix
$\tau$: Random walk mixing time
$r$: Degree assortativity
$c$: Global clustering coefficient
$\oslash$: (Pseudo-) diameter
$S$: Relative size of the largest component
NPs: Node properties
EPs: Edge properties

Name	Nodes	Edges	$\left<k\right>$	$\sigma_k$	$\lambda_h$	$\tau$	$r$	$c$	$\oslash$	$S$	Kind	Mode	NPs	EPs	gt	GraphML	GML	csv
yoochoose	11,616,448	42,406,488	3.65	251.10	295.51	394.98	-0.20	0.00	24	1.00	Directed	Bipartite	is_session id	time buy test category special_offer price quantity	266.8 MiB	450.1 MiB	433.2 MiB	479.2 MiB