Method for the Development of Recommendation Systems, Customizable to Domains, with Deep GRU Network

Arseny Korotaev, Lyudmila Lyadova

2018

Abstract

The GRU-based recurrent neural networks (RNN) for constructing recommendation systems are proposed. Such systems are mainly developed by large companies for specific domains. At the same time, small companies don’t have the necessary resources to develop their own unique systems. Therefore, they need universal recommendation system (or recommender platform) automatically customized for a specific domain. This system allows to develop own recommendation system from scratch for companies whose services are under development. The RNN-based approach is proposed for session-based recommendation with automatically modelling of the domain. This approach is based on the content analysis of the web sites. Several modifications to classic RNNs such as a ranking loss function that make it more viable for this specific problem are considered. General scheme of the approach and architecture of the recommendation system based on proposed scheme are described in this paper.

Paper Citation

in Harvard Style

Korotaev A. and Lyadova L. (2018). Method for the Development of Recommendation Systems, Customizable to Domains, with Deep GRU Network. In Proceedings of the 10th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2018) - Volume 2: KEOD; ISBN 978-989-758-330-8, SciTePress, pages 231-236. DOI: 10.5220/0006933302310236

in Bibtex Style

@conference{keod18,
author={Arseny Korotaev and Lyudmila Lyadova},
title={Method for the Development of Recommendation Systems, Customizable to Domains, with Deep GRU Network},
booktitle={Proceedings of the 10th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2018) - Volume 2: KEOD},
year={2018},
pages={231-236},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006933302310236},
isbn={978-989-758-330-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 10th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2018) - Volume 2: KEOD
TI - Method for the Development of Recommendation Systems, Customizable to Domains, with Deep GRU Network
SN - 978-989-758-330-8
AU - Korotaev A.
AU - Lyadova L.
PY - 2018
SP - 231
EP - 236
DO - 10.5220/0006933302310236
PB - SciTePress