Convolutional Neural Networks (CNNs) have been recently introduced
in the domain of session-based next item recommendation.
An ordered collection of past items the user has interacted with in
a session (or sequence) are embedded into a 2-dimensional latent
matrix, and treated as an image. The convolution and pooling operations
are then applied to the mapped item embeddings. In this paper,
we first examine the typical session-based CNN recommender and
show that both the generative model and network architecture are
suboptimal when modeling long-range dependencies in the item
sequence. To address the issues, we introduce a simple, but very
effective generative model that is capable of learning high-level
representation from both short- and long-range item dependencies.
The network architecture of the proposed model is formed of a stack
of holed convolutional layers, which can efficiently increase the
receptive fields without relying on the pooling operation. Another
contribution is the effective use of residual block structure in recommender
systems, which can ease the optimization for much deeper
networks. The proposed generative model attains state-of-the-art
accuracy with less training time in the next item recommendation
task. It accordingly can be used as a powerful recommendation
baseline to beat in future, especially when there are long sequences
of user feedback.
%0 Conference Paper
%1 conf/wsdm/YuanKAJ019
%A Yuan, Fajie
%A Karatzoglou, Alexandros
%A Arapakis, Ioannis
%A Jose, Joemon M.
%A He, Xiangnan
%B WSDM
%D 2019
%E Culpepper, J. Shane
%E Moffat, Alistair
%E Bennett, Paul N.
%E Lerman, Kristina
%I ACM
%K thema:personalized_top-n
%P 582-590
%T A Simple Convolutional Generative Network for Next Item Recommendation.
%U http://dblp.uni-trier.de/db/conf/wsdm/wsdm2019.html#YuanKAJ019
%X Convolutional Neural Networks (CNNs) have been recently introduced
in the domain of session-based next item recommendation.
An ordered collection of past items the user has interacted with in
a session (or sequence) are embedded into a 2-dimensional latent
matrix, and treated as an image. The convolution and pooling operations
are then applied to the mapped item embeddings. In this paper,
we first examine the typical session-based CNN recommender and
show that both the generative model and network architecture are
suboptimal when modeling long-range dependencies in the item
sequence. To address the issues, we introduce a simple, but very
effective generative model that is capable of learning high-level
representation from both short- and long-range item dependencies.
The network architecture of the proposed model is formed of a stack
of holed convolutional layers, which can efficiently increase the
receptive fields without relying on the pooling operation. Another
contribution is the effective use of residual block structure in recommender
systems, which can ease the optimization for much deeper
networks. The proposed generative model attains state-of-the-art
accuracy with less training time in the next item recommendation
task. It accordingly can be used as a powerful recommendation
baseline to beat in future, especially when there are long sequences
of user feedback.
%@ 978-1-4503-5940-5
@inproceedings{conf/wsdm/YuanKAJ019,
abstract = {Convolutional Neural Networks (CNNs) have been recently introduced
in the domain of session-based next item recommendation.
An ordered collection of past items the user has interacted with in
a session (or sequence) are embedded into a 2-dimensional latent
matrix, and treated as an image. The convolution and pooling operations
are then applied to the mapped item embeddings. In this paper,
we first examine the typical session-based CNN recommender and
show that both the generative model and network architecture are
suboptimal when modeling long-range dependencies in the item
sequence. To address the issues, we introduce a simple, but very
effective generative model that is capable of learning high-level
representation from both short- and long-range item dependencies.
The network architecture of the proposed model is formed of a stack
of holed convolutional layers, which can efficiently increase the
receptive fields without relying on the pooling operation. Another
contribution is the effective use of residual block structure in recommender
systems, which can ease the optimization for much deeper
networks. The proposed generative model attains state-of-the-art
accuracy with less training time in the next item recommendation
task. It accordingly can be used as a powerful recommendation
baseline to beat in future, especially when there are long sequences
of user feedback.},
added-at = {2020-12-13T20:45:24.000+0100},
author = {Yuan, Fajie and Karatzoglou, Alexandros and Arapakis, Ioannis and Jose, Joemon M. and He, Xiangnan},
biburl = {https://www.bibsonomy.org/bibtex/2002c2eac8825f4304e27f3fe353929ad/helenaf},
booktitle = {WSDM},
crossref = {conf/wsdm/2019},
editor = {Culpepper, J. Shane and Moffat, Alistair and Bennett, Paul N. and Lerman, Kristina},
ee = {https://doi.org/10.1145/3289600.3290975},
interhash = {9085f58bb3e951d528235a2cd4fa7068},
intrahash = {002c2eac8825f4304e27f3fe353929ad},
isbn = {978-1-4503-5940-5},
keywords = {thema:personalized_top-n},
pages = {582-590},
publisher = {ACM},
timestamp = {2020-12-13T20:45:24.000+0100},
title = {A Simple Convolutional Generative Network for Next Item Recommendation.},
url = {http://dblp.uni-trier.de/db/conf/wsdm/wsdm2019.html#YuanKAJ019},
year = 2019
}