Standard

Recurrent knowledge graph embedding for effective recommendation. / Sun, Zhu; Yang, Jie; Zhang, Jie; Bozzon, Alessandro; Huang, Long Kai; Xu, Chi.

RecSys '18 : Proceedings of the 12th ACM Conference on Recommender Systems. New York, NY : Association for Computer Machinery, 2018. p. 297-305.

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review

Harvard

Sun, Z, Yang, J, Zhang, J, Bozzon, A, Huang, LK & Xu, C 2018, Recurrent knowledge graph embedding for effective recommendation. in RecSys '18 : Proceedings of the 12th ACM Conference on Recommender Systems. Association for Computer Machinery, New York, NY, pp. 297-305, 12th ACM Conference on Recommender Systems, RecSys 2018, Vancouver, Canada, 2/10/18. https://doi.org/10.1145/3240323.3240361

APA

Sun, Z., Yang, J., Zhang, J., Bozzon, A., Huang, L. K., & Xu, C. (2018). Recurrent knowledge graph embedding for effective recommendation. In RecSys '18 : Proceedings of the 12th ACM Conference on Recommender Systems (pp. 297-305). New York, NY: Association for Computer Machinery. https://doi.org/10.1145/3240323.3240361

Vancouver

Sun Z, Yang J, Zhang J, Bozzon A, Huang LK, Xu C. Recurrent knowledge graph embedding for effective recommendation. In RecSys '18 : Proceedings of the 12th ACM Conference on Recommender Systems. New York, NY: Association for Computer Machinery. 2018. p. 297-305 https://doi.org/10.1145/3240323.3240361

Author

Sun, Zhu ; Yang, Jie ; Zhang, Jie ; Bozzon, Alessandro ; Huang, Long Kai ; Xu, Chi. / Recurrent knowledge graph embedding for effective recommendation. RecSys '18 : Proceedings of the 12th ACM Conference on Recommender Systems. New York, NY : Association for Computer Machinery, 2018. pp. 297-305

BibTeX

@inproceedings{9a3559e927b647cd820dd7ecc76cbc06,
title = "Recurrent knowledge graph embedding for effective recommendation",
abstract = "Knowledge graphs (KGs) have proven to be effective to improve recommendation. Existing methods mainly rely on hand-engineered features from KGs (e.g., meta paths), which requires domain knowledge. This paper presents RKGE, a KG embedding approach that automatically learns semantic representations of both entities and paths between entities for characterizing user preferences towards items. Specifically, RKGE employs a novel recurrent network architecture that contains a batch of recurrent networks to model the semantics of paths linking a same entity pair, which are seamlessly fused into recommendation. It further employs a pooling operator to discriminate the saliency of different paths in characterizing user preferences towards items. Extensive validation on real-world datasets shows the superiority of RKGE against state-of-the-art methods. Furthermore, we show that RKGE provides meaningful explanations for recommendation results.",
keywords = "Attention Mechanism, Knowledge Graph, Recurrent Neural Network, Semantic Representation",
author = "Zhu Sun and Jie Yang and Jie Zhang and Alessandro Bozzon and Huang, {Long Kai} and Chi Xu",
note = "Accepted author manuscript",
year = "2018",
doi = "10.1145/3240323.3240361",
language = "English",
isbn = "978-1-4503-5901-6",
pages = "297--305",
booktitle = "RecSys '18",
publisher = "Association for Computer Machinery",

}

RIS

TY - GEN

T1 - Recurrent knowledge graph embedding for effective recommendation

AU - Sun, Zhu

AU - Yang, Jie

AU - Zhang, Jie

AU - Bozzon, Alessandro

AU - Huang, Long Kai

AU - Xu, Chi

N1 - Accepted author manuscript

PY - 2018

Y1 - 2018

N2 - Knowledge graphs (KGs) have proven to be effective to improve recommendation. Existing methods mainly rely on hand-engineered features from KGs (e.g., meta paths), which requires domain knowledge. This paper presents RKGE, a KG embedding approach that automatically learns semantic representations of both entities and paths between entities for characterizing user preferences towards items. Specifically, RKGE employs a novel recurrent network architecture that contains a batch of recurrent networks to model the semantics of paths linking a same entity pair, which are seamlessly fused into recommendation. It further employs a pooling operator to discriminate the saliency of different paths in characterizing user preferences towards items. Extensive validation on real-world datasets shows the superiority of RKGE against state-of-the-art methods. Furthermore, we show that RKGE provides meaningful explanations for recommendation results.

AB - Knowledge graphs (KGs) have proven to be effective to improve recommendation. Existing methods mainly rely on hand-engineered features from KGs (e.g., meta paths), which requires domain knowledge. This paper presents RKGE, a KG embedding approach that automatically learns semantic representations of both entities and paths between entities for characterizing user preferences towards items. Specifically, RKGE employs a novel recurrent network architecture that contains a batch of recurrent networks to model the semantics of paths linking a same entity pair, which are seamlessly fused into recommendation. It further employs a pooling operator to discriminate the saliency of different paths in characterizing user preferences towards items. Extensive validation on real-world datasets shows the superiority of RKGE against state-of-the-art methods. Furthermore, we show that RKGE provides meaningful explanations for recommendation results.

KW - Attention Mechanism

KW - Knowledge Graph

KW - Recurrent Neural Network

KW - Semantic Representation

UR - http://www.scopus.com/inward/record.url?scp=85056751122&partnerID=8YFLogxK

U2 - 10.1145/3240323.3240361

DO - 10.1145/3240323.3240361

M3 - Conference contribution

SN - 978-1-4503-5901-6

SP - 297

EP - 305

BT - RecSys '18

PB - Association for Computer Machinery

CY - New York, NY

ER -

ID: 47581040