Yunjia Xi
ABSTRACT
As an emerging field driven by industrial applications, integrated re-ranking combines lists from upstream sources into a single list, and presents it to the user. The quality of integrated re-ranking is especially sensitive to real-time user behaviors and preferences. However, existing methods are all built on the cloud-to-edge framework, where mixed lists are generated by the cloud model and then sent to the devices. Despite its effectiveness, such a framework fails to capture users' real-time preferences due to the network bandwidth and latency. Hence, we propose to place the integrated re-ranking model on devices, allowing for the full exploitation of real-time behaviors. To achieve this, we need to address two key issues: first, how to extract users' preferences for different sources from heterogeneous and imbalanced user behaviors; second, how to explore the correlation between the extracted personalized preferences and the candidate items. In this work, we present the first on-Device Integrated Re-ranking framework, DIR, to avoid delays in processing real-time user behaviors. DIR includes a multi-sequence behavior modeling module to extract the user's source-level preferences, and a preference-adaptive re-ranking module to incorporate personalized source-level preferences into the re-ranking of candidate items. Besides, we design exposure loss and utility loss to jointly optimize exposure fairness and overall utility. Extensive experiments on three datasets show that DIR significantly outperforms the state-of-the-art baselines in utility-based and fairness-based metrics.
Supplemental Material
- 2020. MindSpore. https://www.mindspore.cn/Google Scholar
- Qingyao Ai, Keping Bi, Jiafeng Guo, and W Bruce Croft. 2018. Learning a deep listwise context model for ranking refinement. In The 41st international ACM SIGIR conference on research & development in information retrieval. 135--144.Google ScholarDigital Library
- Javier Alcaraz, Mercedes Landete, and Juan F. Monge. 2022. Rank Aggregation: Models and Algorithms. 153--178.Google Scholar
- Irwan Bello, Sayali Kulkarni, Sagar Jain, Craig Boutilier, Ed Chi, Elad Eban, Xiyang Luo, Alan Mackey, and Ofer Meshi. 2018. Seq2slate: Re-ranking and slate optimization with rnns. arXiv preprint arXiv:1810.02019 (2018).Google Scholar
- Ting Chen, Lala Li, and Yizhou Sun. 2020. Differentiable product quantization for end-to-end embedding compression. In International Conference on Machine Learning. PMLR, 1617--1626.Google Scholar
- Tong Chen, Hongzhi Yin, Yujia Zheng, Zi Huang, Yang Wang, and Meng Wang. 2021b. Learning Elastic Embeddings for Customizing On-Device Recommenders. In Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining. 138--147.Google ScholarDigital Library
- Zeyuan Chen, Jiangchao Yao, Feng Wang, Kunyang Jia, Bo Han, Wei Zhang, and Hongxia Yang. 2021a. MC2-SF: Slow-Fast Learning for Mobile-Cloud Collaborative Recommendation. https://doi.org/10.48550/ARXIV.2109.12314Google Scholar
- Kyunghyun Cho, Bart van Merriënboer, Dzmitry Bahdanau, and Yoshua Bengio. 2014. On the Properties of Neural Machine Translation: Encoder--Decoder Approaches. In Proceedings of SSST-8, Eighth Workshop on Syntax, Semantics and Structure in Statistical Translation. 103--111.Google ScholarCross Ref
- Shangfeng Dai, Haobin Lin, Zhichen Zhao, Jianying Lin, Honghuan Wu, Zhe Wang, Sen Yang, and Ji Liu. 2021. POSO: Personalized Cold Start Modules for Large-scale Recommender Systems. https://doi.org/10.48550/ARXIV.2108.04690Google Scholar
- Zhengxiao Du, Xiaowei Wang, Hongxia Yang, Jingren Zhou, and Jie Tang. 2019. Sequential Scenario-Specific Meta Learner for Online Recommendation. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2895--2904.Google ScholarDigital Library
- Tom Fawcett. 2006. An introduction to ROC analysis. Pattern Recognition Letters, Vol. 27, 8 (2006), 861--874.Google ScholarDigital Library
- Mingsheng Fu, Anubha Agrawal, Athirai A. Irissappane, Jie Zhang, Liwei Huang, and Hong Qu. 2022. Deep Reinforcement Learning Framework for Category-Based Item Recommendation. IEEE Transactions on Cybernetics, Vol. 52, 11 (2022), 12028--12041. https://doi.org/10.1109/TCYB.2021.3089941Google ScholarCross Ref
- Xudong Gong, Qinlin Feng, Yuan Zhang, Jiangling Qin, Weijie Ding, Biao Li, Peng Jiang, and Kun Gai. 2022. Real-Time Short Video Recommendation on Mobile Devices. In Proceedings of the 31st ACM International Conference on Information & Knowledge Management. 3103--3112.Google ScholarDigital Library
- Yu Gong, Ziwen Jiang, Yufei Feng, Binbin Hu, Kaiqi Zhao, Qingwen Liu, and Wenwu Ou. 2020. EdgeRec: Recommender System on Edge in Mobile Taobao. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management. 2477--2484.Google ScholarDigital Library
- Wei Guo, Can Zhang, Zhicheng He, Jiarui Qin, Huifeng Guo, Bo Chen, Ruiming Tang, Xiuqiang He, and Rui Zhang. 2022. MISS: Multi-Interest Self-Supervised Learning Framework for Click-Through Rate Prediction. In 2022 IEEE 38th International Conference on Data Engineering (ICDE). 727--740. https://doi.org/10.1109/ICDE53745.2022.00059Google Scholar
- Sepp Hochreiter and Jürgen Schmidhuber. 1997. Long Short-Term Memory. Neural Computation, Vol. 9 (1997), 1735--1780.Google ScholarDigital Library
- Eric Jang, Shixiang Gu, and Ben Poole. 2016. Categorical Reparameterization with Gumbel-Softmax. https://arxiv.org/abs/1611.01144Google Scholar
- Kalervo Järvelin and Jaana Kekäläinen. 2002. Cumulated Gain-Based Evaluation of IR Techniques. ACM Trans. Inf. Syst. (2002), 422--446.Google Scholar
- James M. Joyce. 2011. Kullback-Leibler Divergence. Springer Berlin Heidelberg, Berlin, Heidelberg, 720--722. https://doi.org/10.1007/978-3-642-04898-2_327Google Scholar
- Jakub Konečný, H. Brendan McMahan, Felix X. Yu, Peter Richtarik, Ananda Theertha Suresh, and Dave Bacon. 2016. Federated Learning: Strategies for Improving Communication Efficiency. In NIPS.Google Scholar
- Yi Li, Jieming Zhu, Weiwen Liu, Liangcai Su, Guohao Cai, Qi Zhang, Ruiming Tang, Xi Xiao, and Xiuqiang He. 2022. PEAR: Personalized Re-ranking with Contextualized Transformer for Recommendation. arXiv preprint arXiv:2203.12267 (2022).Google Scholar
- Guogang Liao, Ze Wang, Xiaoxu Wu, Xiaowen Shi, Chuheng Zhang, Yongkang Wang, Xingxing Wang, and Dong Wang. 2022. Cross DQN: Cross Deep Q Network for Ads Allocation in Feed. In WWW. 401--409.Google Scholar
- J. Lin. 1991. Divergence measures based on the Shannon entropy. IEEE Transactions on Information Theory, Vol. 37, 1 (1991), 145--151.Google ScholarDigital Library
- Weiwen Liu, Qing Liu, Ruiming Tang, Junyang Chen, Xiuqiang He, and Pheng Ann Heng. 2020. Personalized Re-ranking with Item Relationships for E-commerce. In Proceedings of the 29th ACM International Conference on Information & Knowledge Management. 925--934.Google ScholarDigital Library
- Weiwen Liu, Yunjia Xi, Jiarui Qin, Fei Sun, Bo Chen, Weinan Zhang, Rui Zhang, and Ruiming Tang. 2022. Neural Re-ranking in Multi-stage Recommender Systems: A Review. arXiv preprint arXiv:2202.06602 (2022).Google Scholar
- Yan Lu, Yuanchao Shu, Xu Tan, Yunxin Liu, Mengyu Zhou, Qi Chen, and Dan Pei. 2019. Collaborative Learning between Cloud and End Devices: An Empirical Study on Location Prediction. In Proceedings of the 4th ACM/IEEE Symposium on Edge Computing. 139--151.Google ScholarDigital Library
- I. Misra, A. Shrivastava, A. Gupta, and M. Hebert. 2016. Cross-Stitch Networks for Multi-task Learning. In 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). 3994--4003.Google Scholar
- Natwar Modani, Deepali Jain, Ujjawal Soni, Gaurav Kumar Gupta, and Palak Agarwal. 2017. Fairness Aware Recommendations on Behance. In Advances in Knowledge Discovery and Data Mining, Jinho Kim, Kyuseok Shim, Longbing Cao, Jae-Gil Lee, Xuemin Lin, and Yang-Sae Moon (Eds.). 144--155.Google Scholar
- Liang Pang, Jun Xu, Qingyao Ai, Yanyan Lan, Xueqi Cheng, and Jirong Wen. 2020. Setrank: Learning a permutation-invariant ranking model for information retrieval. In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. 499--508.Google ScholarDigital Library
- Changhua Pei, Yi Zhang, Yongfeng Zhang, Fei Sun, Xiao Lin, Hanxiao Sun, Jian Wu, Peng Jiang, Junfeng Ge, Wenwu Ou, et al. 2019. Personalized re-ranking for recommendation. In Proceedings of the 13th ACM conference on recommender systems. 3--11.Google ScholarDigital Library
- Evaggelia Pitoura, Kostas Stefanidis, and Georgia Koutrika. 2021. Fairness in Rankings and Recommendations: An Overview. The VLDB Journal, Vol. 31 (oct 2021), 431--458.Google Scholar
- Divyasheel Sharma and Santonu Sarkar. 2022. Enabling Inference and Training of Deep Learning Models for AI Applications on IoT Edge Devices. Springer International Publishing, Cham, 267--283.Google Scholar
- Weiping Song, Chence Shi, Zhiping Xiao, Zhijian Duan, Yewen Xu, Ming Zhang, and Jian Tang. 2019. AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks. In CIKM. 1161--1170.Google Scholar
- Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, Ł ukasz Kaiser, and Illia Polosukhin. 2017. Attention is All you Need. In NIPS, Vol. 30.Google Scholar
- Qinyong Wang, Hongzhi Yin, Tong Chen, Zi Huang, Hao Wang, Yanchang Zhao, and Nguyen Quoc Viet Hung. 2020. Next Point-of-Interest Recommendation on Resource-Constrained Mobile Devices. In WWW. 906--916.Google Scholar
- Yunjia Xi, Weiwen Liu, Jieming Zhu, Xilong Zhao, Xinyi Dai, Ruiming Tang, Weinan Zhang, Rui Zhang, and Yong Yu. 2022. Multi-Level Interaction Reranking with User Behavior History. In Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval.Google ScholarDigital Library
- Wei Xia, Weiwen Liu, Yifan Liu, and Ruiming Tang. 2022a. Balancing Utility and Exposure Fairness for Integrated Ranking with Reinforcement Learning. In CIKM. 4590--4594.Google Scholar
- Xin Xia, Hongzhi Yin, Junliang Yu, Qinyong Wang, Guandong Xu, and Quoc Viet Hung Nguyen. 2022b. On-Device Next-Item Recommendation with Self-Supervised Knowledge Distillation. In SIGIR. 546--555.Google Scholar
- Xin Xia, Junliang Yu, Qinyong Wang, Chaoqun Yang, Nguyen Quoc Viet Hung, and Hongzhi Yin. 2023. Efficient On-Device Session-Based Recommendation. ACM Trans. Inf. Syst. (jan 2023).Google Scholar
- Ruobing Xie, Shaoliang Zhang, Rui Wang, Feng Xia, and Leyu Lin. 2021. Hierarchical Reinforcement Learning for Integrated Recommendation. In Proceedings of the AAAI Conference on Artificial Intelligence. 4521--4528.Google ScholarCross Ref
- Yikai Yan, Chaoyue Niu, Renjie Gu, Fan Wu, Shaojie Tang, Lifeng Hua, Chengfei Lyu, and Guihai Chen. 2022. On-Device Learning for Model Personalization with Large-Scale Cloud-Coordinated Domain Adaption. In SIGIR. 2180--2190.Google Scholar
- Jiangchao Yao, Feng Wang, Kunyang Jia, Bo Han, Jingren Zhou, and Hongxia Yang. 2021. Device-Cloud Collaborative Learning for Recommendation. In SIGIR. 3865--3874.Google Scholar
- Chunxing Yin, Bilge Acun, Xing Liu, and Carole-Jean Wu. 2021. TT-Rec: Tensor Train Compression for Deep Learning Recommendation Models. In Proceedings of the 4th MLSys Conference.Google Scholar
- Guorui Zhou, Xiaoqiang Zhu, Chenru Song, Ying Fan, Han Zhu, Xiao Ma, Yanghui Yan, Junqi Jin, Han Li, and Kun Gai. 2018. Deep Interest Network for Click-Through Rate Prediction. In CIKM. 1059--1068.Google Scholar
Index Terms
- On-device Integrated Re-ranking with Heterogeneous Behavior Modeling
Recommendations
A smart-device news recommendation technology based on the user click behavior
EDB '16: Proceedings of the Sixth International Conference on Emerging Databases: Technologies, Applications, and TheoryStudies have been conducted in regard to personalized news recommendation using collaborative filtering mechanisms based on users' click behaviors. However, few existing studies have focused on news recommendations depending on the rates of news-...
Adaptive recommendation for preferred information and browsing action based on web-browsing behavior
DASFAA'10: Proceedings of the 15th international conference on Database Systems for Advanced Applications - Volume Part IIA Web recommender system based on the inference from a user’s Web-browsing behavior has been proposed and implemented. This system is capable of recommending items of interest to a user and specific Web-browsing action on the current item using a novel ...
An Improved Collaborative Filtering Approach Based on User Ranking and Item Clustering
IDCS 2013: Proceedings of the 6th International Conference on Internet and Distributed Computing Systems - Volume 8223Collaborative filtering is one of the most successful technologies applied in recommender systems in multiple domains. With the increasing growth of users and items involved in recommender systems, some inherent weaknesses of traditional collaborating ...
Access Granted
Comments