Knowledge-Aware Recommendation Algorithm Combined with Attention Mechanism

doi:10.3778/j.issn.1002-8331.2011-0054

Abstract

Abstract: The application of knowledge graphs in recommender systems has attracted more and more attention, which can effectively solve the data sparsity and cold start problems in recommender systems. However, when the existing path-based and embedded-based knowledge-aware recommendation algorithms merge entities in the knowledge graph to represent users, they do not consider that the importance of entities to users is not the same, and the recommendation results will be affected by unrelated entities. Aiming at the limitations of the existing methods, a new knowledge-aware recommendation algorithm combined with the attention mechanism is proposed, and an end-to-end framework for incorporating the knowledge graph into the recommendation system is given. From the user’s historical click items, multiple entity sets are expanded on the knowledge graph, and the user’s preference distribution is calculated through the attention mechanism, and the final click probability is predicted accordingly. Through comparative experiments with traditional recommendation algorithms on two real public data sets, the results show that this method has achieved significant improvement under the evaluation of multiple common indicators（such as AUC, ACC and Recall@top-K）.

Key words: recommendation system, knowledge graph, attention mechanism, entity communication

摘要： 知识图谱在推荐系统中的应用越来越受重视，可以有效地解决推荐系统中存在的数据稀疏性和冷启动问题。但现有的基于路径和基于嵌入的知识感知推荐算法在合并知识图谱中的实体来表示用户时，并没有考虑到实体对于用户的重要性并不相同，推荐结果会受到无关实体的影响。针对现有方法的局限性，提出了一种新的结合注意力机制的知识感知推荐算法，并给出一种将知识图谱合并到推荐系统中的端到端框架。由用户的历史点击项在知识图谱上扩展出多个实体集，通过注意力机制来计算用户的偏好分布，据此预测最终的点击概率。通过在两个真实的公共数据集上与传统的推荐算法进行对比实验，结果表明，该方法在多个通用指标（例如AUC、ACC和Recall@top-K）的评估下均取得了明显提升。

关键词: 推荐系统, 知识图谱, 注意力机制, 实体传播

ZHANG Xin, LIU Siyuan, XU Yanling. Knowledge-Aware Recommendation Algorithm Combined with Attention Mechanism[J]. Computer Engineering and Applications, 2022, 58(9): 168-174.

张昕, 刘思远, 徐雁翎. 结合注意力机制的知识感知推荐算法[J]. 计算机工程与应用, 2022, 58(9): 168-174.

References

[1] KOREN Y，BELL R，VOLINSKY C.Matrix factorization techniques for recommender systems[J].Computer，2009，42（8）：30-37.
[2] WANG H，ZHANG F，HOU M，et al.Shine：signed hetero-geneous information network embedding for sentiment l?ink prediction[C]//Proceedings of the Eleventh ACM International Conference on Web Search and Data Mining，2018：592-600.
[3] WANG Q，MAO Z，WANG B，et al.Knowledge graph embedding：a survey of approaches and applications[J].IEEE Transactions on Knowledge and Data Engineering，2017，29（12）：2724-2743.
[4] LIN H，LIU Y，WANG W，et al.Learning entity and relation embeddings for knowledge resolution[J].Procedia Computer Science，2017，108：345-354.
[5] XU C，BAI Y，BIAN J，et al.Rc-net：a general framework for incorporating knowledge into word representations[C]//Proceedings of the 23rd ACM International Conference on Information and Knowledge Management，2014：1219-1228.
[6] 秦川，祝恒书，庄福振，等.基于知识图谱的推荐系统研究综述[J].中国科学：信息科学，2020，50（7）：937-956.
QIN C，ZU H S，ZHUANG F Z，et al.A survey on knowledge graph-based recommender systems[J].Scientia Sinica：Informationis，2020，50（7）：937-956.
[7] ZHANG F，YUAN N J，LIAN D，et al.Collaborative knowledge base embedding for recommender systems[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining，2016：353-362.
[8] WANG H W，ZHANG F Z，XIE X，et al.DKN：Deep knowledge-aware network for news recommendation[C]//The 27th International Conference on World Wide Web，2018：1835-1844.
[9] 李浩，张亚钏，康雁，等.融合循环知识图谱和协同过滤电影推荐算法[J].计算机工程与应用，2020，56（2）：106-114.
LI H，ZHANG Y C，KANG Y，et al.Fusion recurrent knowledge graph and collaborative filtering movie recommendation algorithm[J].Computer Engineering and Applications，2020，56（2）：106-114.
[10] YU X，REN X，SUN Y Z，et al.Personalized entity recommendation：a heterogeneous information network approach[C]//Proceedings of the 7th ACM International Conference on Web Search and Data Mining，2014：283-292.
[11] LUO C，PANG W，WANG Z，et al.Hetecf：social-based collaborative filtering recommendation using heterogeneous relations[C]//2014 IEEE International Conference on Data Mining，2014：917-922.
[12] BORDES A，USUNIER N，GARCIA-DURAN A，et al.Translating embeddings for modeling multi-relationaldata[C]//Advances in Neural Information Processing Systems，2013：2787-2795.
[13] WANG Z，ZHANG J，FENG J，et al.Knowledge graph embedding by translating on hyperplanes[C]//Proceedings of the Twenty-Eighth AAAI Conference on Artificial Intelligence，2014：1112-1119.
[14] TROUILLON T，WELBL J，RIEDEL S，et al.Complex embeddings for simple link prediction[C]//International Conference on Machine Learning，2016：2071-2080.
[15] Nickel M，Tresp V，Kriegel H P.A three-way model for collective learning on multi-relational data[C]//International Conference on Machine Learning，2011：809-816.
[16] BAHDANAU D，CHO K，BENGIO Y.Neural machine translation by jointly learning toalign and translate[J].arXiv：1409.0473，2014.
[17] 谭台哲，晏家斌.基于注意力模型的混合推荐系统[J].计算机工程与应用，2020，56（13）：172-180.
TAN T Z，YAN J B.Hybrid recommendation system based on self-attention model[J].Computer Engineering and Applications，2020，56（13）：172-180.
[18] MIKOLOV T，SUTSKEVER I，CHEN K，et al.Distributed representations of words and phrases and their compositionality[C]//Proceedings of the 26th International Conference on Neural Information Processing Systems，2013：3111-3119.
[19] WANG H W，ZHANG F Z，ZHAO M，et al.Multi-task feature learning for knowledge graph enhanced recommendation[J].arXiv：1901.08907，2019.
[20] RENDLE S.Factorization machines with libfm[J].ACM Transactions on Intelligent Systems and Technology（TIST），2012，3（3）：1-22.
[21] CHENG H T，KOC L，HARMSEN J，et al.Wide & deep learning for recommender systems[C]//Proceedings of the 1st Workshop on Deep Learning for Recommender Systems，2016：7-10.