TY - GEN
T1 - Hierarchical Graph Neural Network with Cross-Attention for Cross-Device User Matching
AU - Taghibakhshi, Ali
AU - Ma, Mingyuan
AU - Aithal, Ashwath
AU - Yilmaz, Onur
AU - Maron, Haggai
AU - West, Matthew
N1 - Acknowledgements. This work was partially supported by Research Institute of Konan University, and by JSPS KAKENHI 20K12085 and 19H04218.
Acknowledgments. This research is supported by the European Union’s Horizon 2020 research and innovation program under grant agreement No 875171, project QUALITOP (Monitoring multidimensional aspects of QUAlity of Life after cancer ImmunoTherapy - an Open smart digital Platform for personalized prevention and patient management).
Acknowledgements. This work is funded by the French National Research Agency under grant ANR-20-CE23-0002.
Acknowledgement. This work is partially supported by Natural Sciences and Engineering Research Council of Canada (NSERC) and University of Manitoba.
Acknowledgements. This research/project is supported by the National Research Foundation, Singapore under its Industry Alignment Fund - Pre-positioning (IAF-PP) Funding Initiative. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not reflect the views of National Research Foundation, Singapore. Additionally, we thank Jiatong Han for helping with the creation of tutorials.
Acknowledgments. This project has been supported in part by funding from GEO Directorate under NSF awards #2204363, #2240022, and #2301397 and the CISE Directorate under NSF award #2305781.
Acknowledgments. The authors would like to thank the French National Centre for Scientific Research (CNRS) for their financial support through the DSCA project FDMI-AMG.
Acknowledgments. This work was supported by the French Gov. in the framework of the Territoire d’Innovation program, an action of the Grand Plan d’Investissement backed by France 2030, Toulouse Métropole and the GIS neOCampus.
Acknowledgements. This work is supported in part under two grants from NSF (Award #2104004) and NASA (SWR2O2R Grant #80NSSC22K0272).
This research was supported by NVIDIA Corporation.
Acknowledgement. This work was supported by the Research Center Trustworthy Data Science and Security, an institution of the University Alliance Ruhr.
This work was supported by the Federal Ministry of Education and Research (BMBF) under Grand No. 16DHB4020.
PY - 2023
Y1 - 2023
N2 - Cross-device user matching is a critical problem in numerous domains, including advertising, recommender systems, and cybersecurity. It involves identifying and linking different devices belonging to the same person, utilizing sequence logs. Previous data mining techniques have struggled to address the long-range dependencies and higher-order connections between the logs. Recently, researchers have modeled this problem as a graph problem and proposed a two-tier graph contextual embedding (TGCE) neural network architecture, which outperforms previous methods. In this paper, we propose a novel hierarchical graph neural network architecture (HGNN), which has a more computationally efficient second level design than TGCE. Furthermore, we introduce a cross-attention (Cross-Att) mechanism in our model, which improves performance by 5% compared to the state-of-the-art TGCE method.
AB - Cross-device user matching is a critical problem in numerous domains, including advertising, recommender systems, and cybersecurity. It involves identifying and linking different devices belonging to the same person, utilizing sequence logs. Previous data mining techniques have struggled to address the long-range dependencies and higher-order connections between the logs. Recently, researchers have modeled this problem as a graph problem and proposed a two-tier graph contextual embedding (TGCE) neural network architecture, which outperforms previous methods. In this paper, we propose a novel hierarchical graph neural network architecture (HGNN), which has a more computationally efficient second level design than TGCE. Furthermore, we introduce a cross-attention (Cross-Att) mechanism in our model, which improves performance by 5% compared to the state-of-the-art TGCE method.
KW - Cross-attention
KW - Graph neural network
KW - User matching
UR - http://www.scopus.com/inward/record.url?scp=85172409302&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85172409302&partnerID=8YFLogxK
U2 - 10.1007/978-3-031-39831-5_28
DO - 10.1007/978-3-031-39831-5_28
M3 - Conference contribution
AN - SCOPUS:85172409302
SN - 9783031398308
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 303
EP - 315
BT - Big Data Analytics and Knowledge Discovery - 25th International Conference, DaWaK 2023, Proceedings
A2 - Wrembel, Robert
A2 - Gamper, Johann
A2 - Kotsis, Gabriele
A2 - Khalil, Ismail
A2 - Tjoa, A Min
PB - Springer
T2 - Big Data Analytics and Knowledge Discovery - 25th International Conference, DaWaK 2023, Proceedings
Y2 - 28 August 2023 through 30 August 2023
ER -