A User Recommendation System Based on Graph Neural Network and Contextual Behavior
The result's identifiers
Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F60076658%3A12510%2F23%3A43906985" target="_blank" >RIV/60076658:12510/23:43906985 - isvavai.cz</a>
Result on the web
<a href="https://mme2023.vse.cz/mme_2023_proceedings.pdf" target="_blank" >https://mme2023.vse.cz/mme_2023_proceedings.pdf</a>
DOI - Digital Object Identifier
—
Alternative languages
Result language
angličtina
Original language name
A User Recommendation System Based on Graph Neural Network and Contextual Behavior
Original language description
Today, recommendation systems are an integral part of e-commerce services on the Internet. In connection with their development, neural networks have become the most used approach to recommender systems. In our post, we will demonstrate the use of graph neural networks to create a recommender system. E-commerce systems can be modeled using a bipartite interaction graph. There are two essential parts to this chart, users and items. In our model, context is added to them and integrated into the mentioned parts of the bipartite graph using the theory of hypothetical functions. Different elements of a bipartite graph can interact using edges. Therefore, modeling the interaction of elements can be transformed into modeling the interaction of nodes on the corresponding graph. We implemented a recommender system model in Python and used relevant libraries, which we tested on standard datasets. These experiments showed the good ability of our model for recommendations. We used the root mean square error (RMSE) and mean absolute error (MAE) indicators.
Czech name
—
Czech description
—
Classification
Type
D - Article in proceedings
CEP classification
—
OECD FORD branch
10102 - Applied mathematics
Result continuities
Project
—
Continuities
I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace
Others
Publication year
2023
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů
Data specific for result type
Article name in the collection
Proceedings of the 41st International Conference on Mathematical Methods in Economics
ISBN
978-80-11-04132-8
ISSN
2788-3965
e-ISSN
—
Number of pages
6
Pages from-to
111-116
Publisher name
Czech Society for Operations Research
Place of publication
Praha
Event location
Praha
Event date
Sep 13, 2023
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—