Evaluating a Session-based Recommender System using Prod2vec in a Commercial Application

Hasan Tercan, Christian Bitter, Todd Bodnar, Philipp Meisen, Tobias Meisen

2021

Abstract

Recommender systems are a central component of many online stores and product websites. An essential functionality of them is to show users new products that they do not yet know they want to buy. Since the users of the website are often unknown to the system, a product recommendation must be made using the current activities within a browser session. In this paper we address this issue in a deep learning-based product-to-product recommendation problem for a commercial website with millions of user interactions. Our proposed approach is based on a prod2vec method for product embeddings, thus recommending those products that often occur together with the target product. Following the idea of word2vec methods from the NLP domain, we train an artificial neural network on user activity data extracted from historical browser sessions. As part of several real A/B tests on the website, we prove that our approach delivers successful product recommendations and outperforms the current system in use. In addition, the results show that the performance can be significantly improved by an appropriate selection of the training data and the time range of historical user interactions.

Paper Citation

in Harvard Style

Tercan H., Bitter C., Bodnar T., Meisen P. and Meisen T. (2021). Evaluating a Session-based Recommender System using Prod2vec in a Commercial Application. In Proceedings of the 23rd International Conference on Enterprise Information Systems - Volume 1: ICEIS, ISBN 978-989-758-509-8, pages 610-617. DOI: 10.5220/0010400706100617

in Bibtex Style

@conference{iceis21,
author={Hasan Tercan and Christian Bitter and Todd Bodnar and Philipp Meisen and Tobias Meisen},
title={Evaluating a Session-based Recommender System using Prod2vec in a Commercial Application},
booktitle={Proceedings of the 23rd International Conference on Enterprise Information Systems - Volume 1: ICEIS,},
year={2021},
pages={610-617},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010400706100617},
isbn={978-989-758-509-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 23rd International Conference on Enterprise Information Systems - Volume 1: ICEIS,
TI - Evaluating a Session-based Recommender System using Prod2vec in a Commercial Application
SN - 978-989-758-509-8
AU - Tercan H.
AU - Bitter C.
AU - Bodnar T.
AU - Meisen P.
AU - Meisen T.
PY - 2021
SP - 610
EP - 617
DO - 10.5220/0010400706100617