ABSTRACT
Recommender Systems have been proven to be of great business value in e-commerce platforms and recommendation algorithms play an important role in it. E-commerce platforms provide entrances for customers to enter channels that can meet their specific shopping needs. Trigger items displayed on entrance icons are useful to attract more entering. User instant interest can be explicitly induced by the trigger items and follow-up related target items are recommended accordingly, this recommendation scenario is called trigger-induced recommendation. However, existing trigger-induced recommendation algorithms usually optimize a single task (e.g., click-through rate prediction) based on users’ historical click sequences, and do not consider modeling the similarity between the trigger item and the target item. In this paper, we propose a novel recommendation method named Deep Multifaceted Highlight Network (DMHN) that can model users’ multiple types of behavior sequences simultaneously for multi-objective ranking in trigger-induced recommendation. DMHN utilizes the Multi-Interest Module to model user’s various behavior sequences to capture the different latent user interests, the Similarity Module to focus on the trigger item by modeling the similarity between the target item and the trigger item, and the Progressive Layered Extraction to optimize multiple objectives. Experiments on JD.com real production dataset demonstrate the superiority of DMHN over state-of-the-art methods.
- [1]Guo, Jian, et al. "Gluoncv and gluonnlp: Deep learning in computer vision and natural language processing." The Journal of Machine Learning Research 21.1 (2020): 845-851.Google ScholarDigital Library
- [2]Ioannidou, Anastasia, et al. "Deep learning advances in computer vision with 3d data: A survey." ACM computing surveys (CSUR) 50.2 (2017): 1-38.Google ScholarDigital Library
- [3]Friedman, Jerome H. "Greedy function approximation: a gradient boosting machine." Annals of statistics (2001): 1189-1232.Google ScholarCross Ref
- [4]Chen, Tianqi, and Carlos Guestrin. "Xgboost: A scalable tree boosting system." Proceedings of the 22nd acm sigkdd international conference on knowledge discovery and data mining. 2016.Google Scholar
- [5]Qu, Yanru, et al. "Product-based neural networks for user response prediction." 2016 IEEE 16th international conference on data mining (ICDM). IEEE, 2016.Google Scholar
- [6]Guo, Huifeng, et al. "DeepFM: a factorization-machine based neural network for CTR prediction." arXiv preprint arXiv:1703.04247 (2017).Google Scholar
- [7]Zhou, Guorui, et al. "Deep interest network for click-through rate prediction." Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining. 2018.Google Scholar
- [8]Liu, Jin, et al. "Attention-based BiGRU-CNN for Chinese question classification." Journal of Ambient Intelligence and Humanized Computing (2019): 1-12.Google Scholar
- [9]Tang, Hongyan, et al. "Progressive layered extraction (ple): A novel multi-task learning (mtl) model for personalized recommendations." Proceedings of the 14th ACM Conference on Recommender Systems. 2020.Google Scholar
- [10]Cheng, Heng-Tze, et al. "Wide & deep learning for recommender systems." Proceedings of the 1st workshop on deep learning for recommender systems. 2016.Google Scholar
- [11]Xue, Hong-Jian, et al. "Deep matrix factorization models for recommender systems." IJCAI. Vol. 17. 2017.Google Scholar
- [12]Zhou, Guorui, et al. "Deep interest evolution network for click-through rate prediction." Proceedings of the AAAI conference on artificial intelligence. Vol. 33. No. 01. 2019.Google Scholar
- [13]Gu, Yulong, et al. "Deep multifaceted transformers for multi-objective ranking in large-scale e-commerce recommender systems." Proceedings of the 29th ACM International Conference on Information & Knowledge Management. 2020.Google Scholar
- [14]Chen, Qiwei, et al. "Behavior sequence transformer for e-commerce recommendation in alibaba." Proceedings of the 1st International Workshop on Deep Learning Practice for High-Dimensional Sparse Data. 2019.Google Scholar
- [15]Feng, Yufei, et al. "Deep session interest network for click-through rate prediction." arXiv preprint arXiv:1905.06482 (2019).Google Scholar
- [16]Li, Jiacheng, Yujie Wang, and Julian McAuley. "Time interval aware self-attention for sequential recommendation." Proceedings of the 13th international conference on web search and data mining. 2020.Google Scholar
- [17]Lu, Yichao, Ruihai Dong, and Barry Smyth. "Why I like it: multi-task learning for recommendation and explanation." Proceedings of the 12th ACM Conference on Recommender Systems. 2018.Google Scholar
- [18]Wang, Jialei, et al. "Online multi-task collaborative filtering for on-the-fly recommender systems." Proceedings of the 7th ACM conference on Recommender systems. 2013.Google Scholar
- [19]Ma, Xiao, et al. "Entire space multi-task model: An effective approach for estimating post-click conversion rate." The 41st International ACM SIGIR Conference on Research & Development in Information Retrieval. 2018.Google ScholarDigital Library
- [20]Ma, Jiaqi, et al. "Modeling task relationships in multi-task learning with multi-gate mixture-of-experts." Proceedings of the 24th ACM SIGKDD international conference on knowledge discovery & data mining. 2018.Google Scholar
- [21]Xie, Ruobing, et al. "Real-time relevant recommendation suggestion." Proceedings of the 14th ACM International Conference on Web Search and Data Mining. 2021.Google Scholar
- [22]Shen, Qijie, et al. "Deep interest highlight network for click-through rate prediction in trigger-induced recommendation." Proceedings of the ACM Web Conference 2022. 2022.Google Scholar
- [23]Xia, Yaxian, et al. "Deep Intention-Aware Network for Click-Through Rate Prediction." arXiv preprint arXiv:2211.08650 (2022).Google Scholar
Index Terms
- Deep Multifaceted Highlight Network for Multi-objective Ranking in Trigger-Induced Recommendation
Recommendations
Deep Interest Highlight Network for Click-Through Rate Prediction in Trigger-Induced Recommendation
WWW '22: Proceedings of the ACM Web Conference 2022In many classical e-commerce platforms, personalized recommendation has been proven to be of great business value, which can improve user satisfaction and increase the revenue of platforms. In this paper, we present a new recommendation problem, Trigger-...
Deep Multifaceted Transformers for Multi-objective Ranking in Large-Scale E-commerce Recommender Systems
CIKM '20: Proceedings of the 29th ACM International Conference on Information & Knowledge ManagementRecommender Systems have been playing essential roles in e-commerce portals. Existing recommendation algorithms usually learn the ranking scores of items by optimizing a single task (e.g. Click-through rate prediction) based on users' historical click ...
Deep Evolutional Instant Interest Network for CTR Prediction in Trigger-Induced Recommendation
WSDM '24: Proceedings of the 17th ACM International Conference on Web Search and Data MiningThe recommendation has been playing a key role in many industries, e.g., e-commerce, streaming media, social media, etc. Recently, a new recommendation scenario, called Trigger-Induced Recommendation (TIR), where users are able to explicitly express ...
Comments