Enhancing Conversational Recommendation Systems with Representation Fusion

Abstract

Conversational Recommendation Systems (CRSs) aim to improve recommendation performance by utilizing information from a conversation session. A CRS first constructs questions and then asks users for their feedback in each conversation session to refine better recommendation lists to users. The key design of CRS is to construct proper questions and obtain users’ feedback in response to these questions so as to effectively capture user preferences. Many CRS works have been proposed; however, they suffer from defects when constructing questions for users to answer: (1) employing a dialogue policy agent for constructing questions is one of the most common choices in CRS, but it needs to be trained with a huge corpus, and (2) it is not appropriate that constructing questions from a single policy (e.g., a CRS only selects attributes that the user has interacted with) for all users with different preferences. To address these defects, we propose a novel CRS model, namely a Representation Fusion–based Conversational Recommendation model, where the whole conversation session is divided into two subsessions (i.e., Local Question Search subsession and Global Question Search subsession) and two different question search methods are proposed to construct questions in the corresponding subsessions without employing policy agents. In particular, in the Local Question Search subsession we adopt a novel graph mining method to find questions, where the paths in the graph between users and attributes can eliminate irrelevant attributes; in the Global Question Search subsession we propose to initialize user preference on items with the user and all item historical rating records and construct questions based on user’s preference. Then, we update the embeddings independently over the two subsessions according to user’s feedback and fuse the final embeddings from the two subsessions for the recommendation. Experiments on three real-world recommendation datasets demonstrate that our proposed method outperforms five state-of-the-art baselines.

REFERENCES

1. [1] Aliannejadi Mohammad, Zamani Hamed, Crestani Fabio, and Croft W. Bruce. 2019. Asking clarifying questions in open-domain information-seeking conversations. In Proceedings of the 42nd International ACM Sigir Conference on Research and Development in Information Retrieval. 475–484.
2. [2] Alves Rodrigo, Ledent Antoine, and Kloft Marius. 2021. Burst-induced multi-armed bandit for learning recommendation. In Proceedings of the 15th ACM Conference on Recommender Systems. 292–301.
3. [3] Anthony Martin, Bartlett Peter L., Bartlett Peter L., et al. 1999. Neural Network Learning: Theoretical Foundations. Vol. 9, Cambridge University Press, Cambridge.
4. [4] Bell Robert, Koren Yehuda, and Volinsky Chris. 2007. Modeling relationships at multiple scales to improve accuracy of large recommender systems. In Proceedings of the 13th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 95–104.
5. [5] Bi Keping, Ai Qingyao, Zhang Yongfeng, and Croft Bruce W.. 2019. Conversational product search based on negative feedback. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management. 359–368.
6. [6] Chen Guanzheng, Fang Jinyuan, Meng Zaiqiao, Zhang Qiang, and Liang Shangsong. 2022. Multi-Relational graph representation learning with bayesian gaussian process network. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 36. 5530–5538.
7. [7] Chen Qibin, Lin Junyang, Zhang Yichang, Ding Ming, Cen Yukuo, Yang Hongxia, and Tang Jie. 2019. Towards knowledge-based recommender dialog system. In Proceedings of the Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP’19). 1803–1813.
8. [8] Christakopoulou Konstantina, Beutel Alex, Li Rui, Jain Sagar, and Chi H. Ed. 2018. Q&R: A two-stage approach toward interactive recommendation. In Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 139–148.
9. [9] Christakopoulou Konstantina, Radlinski Filip, and Hofmann Katja. 2016. Towards conversational recommender systems. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 815–824.
10. [10] Gao Chongming, Lei Wenqiang, He Xiangnan, Rijke de Maarten, and Chua Tat-Seng. 2021. Advances and challenges in conversational recommender systems: A survey. AI Open 2 (2021), 100–126.
11. [11] Ge Yingqiang, Zhao Shuya, Zhou Honglu, Pei Changhua, Sun Fei, Ou Wenwu, and Zhang Yongfeng. 2020. Understanding echo chambers in e-commerce recommender systems. In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. 2261–2270.
12. [12] Gentile Claudio, Li Shuai, Kar Purushottam, Karatzoglou Alexandros, Zappella Giovanni, and Etrue Evans. 2017. On context-dependent clustering of bandits. In International Conference on machine learning, PMLR, 1253–1262.
13. [13] Gu Yulong, Ding Zhuoye, Wang Shuaiqiang, and Yin Dawei. 2020. Hierarchical user profiling for e-commerce recommender systems. In Proceedings of the 13th International Conference on Web Search and Data Mining. 223–231.
14. [14] Harper Maxwell F. and Konstan Joseph A.. 2015. The movielens datasets: History and context. ACM Transactions on Interactive Intelligent Systems (TIIS) 5, 4 (2015), 1–19.
15. [15] Hauff Claudia, Kiseleva Julia, Sanderson Mark, Zamani Hamed, and Zhang Yongfeng. 2021. Conversational search and recommendation: Introduction to the special issue. ACM Transactions on Information Systems 39, 4 (2021).
16. [16] He Xiangnan and Chua Tat-Seng. 2017. Neural factorization machines for sparse predictive analytics. In Proceedings of the 40th International ACM SIGIR conference on Research and Development in Information Retrieval. 355–364.
17. [17] He Xiangnan, Deng Kuan, Wang Xiang, Li Yan, Zhang Yongdong, and Wang Meng. 2020. Lightgcn: Simplifying and powering graph convolution network for recommendation. In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. 639–648.
18. [18] Huifeng Guo, Ruiming Tang, Yunming Ye, Zhenguo Li, and Xiuqiang He. 2017. DeepFM: A factorization-machine based neural network for CTR prediction. In Proceedings of the 26th International Joint Conference on Artificial Intelligence. 1725–1731.
19. [19] Iovine Andrea, Lops Pasquale, Narducci Fedelucio, Gemmis Marco de, and Semeraro Giovanni. 2021. Improving preference elicitation in a conversational recommender system with active learning strategies. In Annual ACM Symposium on Applied Computing. 1375–1382.
20. [20] Jannach Dietmar, Manzoor Ahtsham, Cai Wanling, and Chen Li. 2021. A survey on conversational recommender systems. ACM Comput. Surv. (CSUR) 54, 5 (2021), 1–36.
21. [21] Järvelin Kalervo and Kekäläinen Jaana. 2002. Cumulated gain-based evaluation of IR techniques. ACM Trans. Inf. Syst. (TOIS) 20, 4 (2002), 422–446.
22. [22] Jaynes Edwin T.. 1982. On the rationale of maximum-entropy methods. Proc. IEEE 70, 9 (1982), 939–952.
23. [23] Jiang Shaojie, Ren Pengjie, Monz Christof, and Rijke Maarten de. 2019. Improving neural response diversity with frequency-aware cross-entropy loss. In The World Wide Web Conference. 2879–2885.
24. [24] Kang Wang-Cheng and McAuley Julian. 2018. Self-attentive sequential recommendation. In IEEE international Conference on Data Mining (ICDM’18). IEEE, 197–206.
25. [25] Kingma Diederik P. and Ba Jimmy. 2015. Adam: A method for stochastic optimization. In ICLR (Poster).
26. [26] Koren Yehuda, Bell Robert, and Volinsky Chris. 2009. Matrix factorization techniques for recommender systems. IEEE Comput. 42, 8 (2009), 30–37.
27. [27] Lei Wenqiang, Gao Chongming, and Rijke Maarten de. 2021. RecSys 2021 tutorial on conversational recommendation: Formulation, methods, and evaluation. In Proceedings of the 15th ACM Conference on Recommender Systems. 842–844.
28. [28] Lei Wenqiang, He Xiangnan, Rijke Maarten de, and Chua Tat-Seng. 2020. Conversational recommendation: Formulation, methods, and evaluation. In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. 2425–2428.
29. [29] Lei Wenqiang, He Xiangnan, Miao Yisong, Wu Qingyun, Hong Richang, Kan Min-Yen, and Chua Tat-Seng. 2020. Estimation-Action-Reflection: Towards deep interaction between conversational and recommender systems. In Proceedings of the 13th International Conference on Web Search and Data Mining. 304–312.
30. [30] Lei Wenqiang, Zhang Gangyi, He Xiangnan, Miao Yisong, Wang Xiang, Chen Liang, and Chua Tat-Seng. 2020. Interactive path reasoning on graph for conversational recommendation. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2073–2083.
31. [31] Li Dongdong, Ren Zhaochun, Ren Pengjie, Chen Zhumin, Fan Miao, Ma Jun, and Rijke Maarten de. 2021. Semi-Supervised variational reasoning for medical dialogue generation. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. 544–554.
32. [32] Li Guohui, Chen Qi, Zheng Bolong, Hung Nguyen Quoc Viet, Zhou Pan, and Liu Guanfeng. 2020. Time-aspect-sentiment recommendation models based on novel similarity measure methods. ACM Trans. Web 14, 2 (2020), 1–26.
33. [33] Li Lihong, Chu Wei, Langford John, and Schapire Robert E.. 2010. A contextual-bandit approach to personalized news article recommendation. In Proceedings of the 19th International Conference on World Wide Web. 661–670.
34. [34] Li Raymond, Kahou Samira Ebrahimi, Schulz Hannes, Michalski Vincent, Charlin Laurent, and Pal Chris. 2018. Towards deep conversational recommendations. Advances in Neural Information Processing Systems 31 (2018).
35. [35] Li Shijun, Lei Wenqiang, Wu Qingyun, He Xiangnan, Jiang Peng, and Chua Tat-Seng. 2021. Seamlessly unifying attributes and items: Conversational recommendation for cold-start users. ACM Trans Inf. Syst. 39, 4 (2021), 1–29.
36. [36] Liang Shangsong, Luo Yupeng, and Meng Zaiqiao. 2021. Profiling users for question answering communities via flow-based constrained co-embedding model. ACM Trans. Inf. Syst. 40, 2 (2021), 1–38.
37. [37] Liang Shangsong, Tang Shaowei, Meng Zaiqiao, and Zhang Qiang. 2022. Cross-temporal snapshot alignment for dynamic networks. IEEE Trans. Knowl. Data Eng. (2022), 1–15.
38. [38] Liu Bang, Zhang Hanlin, Kong Linglong, and Niu Di. 2021. Factorizing historical user actions for next-day purchase prediction. ACM Trans. Web 16, 1 (2021), 1–26.
39. [39] Liu Zeming, Wang Haifeng, Niu Zheng-Yu, Wu Hua, Che Wanxiang, and Liu Ting. 2020. Towards conversational recommendation over multi-type dialogs. In 58th Annual Meeting of the Association-For-Computational-Linguistics (Acl) Conference Location Electr Network. Assoc Computational Linguistics-Acl Location Stroudsburg, 1036–1049.
40. [40] Nowak Robert. 2008. Generalized binary search. Annual Allerton Conference on Communication, Control, and Computing, 568–574.
41. [41] Pan Yaoxin, Liang Shangsong, Ren Jiaxin, Meng Zaiqiao, and Zhang Qiang. 2021. Personalized, sequential, attentive, metric-aware product search. ACM Trans. Inf. Syst. 40, 2 (2021), 1–29.
42. [42] Pan Zhiqiang, Cai Fei, Chen Wanyu, and Chen Honghui. 2021. Graph co-attentive session-based recommendation. ACM Trans. Inf. Syst. 40, 4 (2021), 1–31.
43. [43] Pei Jiahuan, Ren Pengjie, and Rijke Maarten de. 2021. A cooperative memory network for personalized task-oriented dialogue systems with incomplete user profiles. In Proceedings of the Web Conference 2021. 1552–1561.
44. [44] Pramod Dhanya and Bafna Prafulla. 2022. Conversational Recommender systems techniques, tools, acceptance, and adoption: A state of the art review. Expert Syst. Appl. (2022), 1–18.
45. [45] Priyogi Bilih. 2019. Preference elicitation strategy for conversational recommender system. In Proceedings of the 12th ACM International Conference on Web Search and Data Mining. 824–825.
46. [46] Radlinski Filip, Balog Krisztian, Byrne Bill, and Krishnamoorthi Karthik. 2019. Coached conversational preference elicitation: A case study in understanding movie preferences. In Proceedings of the 20th Annual SIGdial Meeting on Discourse and Dialogue. 353–360.
47. [47] Racinsk y Matej. 2018. MyAnimeList dataset. Kaggle (2018).
48. [48] Ren Xuhui, Yin Hongzhi, Chen Tong, Wang Hao, Huang Zi, and Zheng Kai. 2021. Learning to ask appropriate questions in conversational recommendation. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. 808–817.
49. [49] Ren Xuhui, Yin Hongzhi, Chen Tong, Wang Hao, Hung Nguyen Quoc Viet, Huang Zi, and Zhang Xiangliang. 2020. Crsal: Conversational recommender systems with adversarial learning. ACM Trans. Inf. Syst. 38, 4 (2020), 1–40.
50. [50] Rendle Steffen. 2010. Factorization machines. In 2010 IEEE International Conference on Data Mining, IEEE, 995–1000.
51. [51] Schafer J. Ben, Frankowski Dan, Herlocker Jon, and Sen Shilad. 2007. Collaborative filtering recommender systems. 291–324.
52. [52] Sepliarskaia Anna, Kiseleva Julia, Radlinski Filip, and Rijke Maarten de. 2018. Preference elicitation as an optimization problem. In Proceedings of the 12th ACM Conference on Recommender Systems. 172–180.
53. [53] Shani Guy, Heckerman David, Brafman Ronen I., and Boutilier Craig. 2005. An MDP-based recommender system.Journal of Machine Learning Research 6, 9 (2005).
54. [54] Sun Yueming and Zhang Yi. 2018. Conversational recommender system. In Proceeding of the 41st International ACM SIGIR Conference on Research & Development in iInformation Retrieval. 235–244.
55. [55] Wang Huazheng, Wu Qingyun, and Wang Hongning. 2017. Factorization bandits for interactive recommendation. In Thirty-first AAAI Conference on Artificial Intelligence.
56. [56] Wang Xiang, He Xiangnan, Wang Meng, Feng Fuli, and Chua Tat-Seng. 2019. Neural graph collaborative filtering. In Proceedings of the 42nd international ACM SIGIR Conference on Research and Development in Information Retrieval. 165–174.
57. [57] Wu Bin, Meng Zaiqiao, Zhang Qiang, and Liang Shangsong. 2022. Meta-Learning helps personalized product search. In Proceedings of the ACM Web Conference 2022. 2277–2287.
58. [58] Wu Jiancan, Wang Xiang, Feng Fuli, He Xiangnan, Chen Liang, Lian Jianxun, and Xie Xing. 2021. Self-supervised graph learning for recommendation. In Proceedings of the 44th International ACM SIGIR Conference on Research and Development in Information Retrieval. 726–735.
59. [59] Wu Qingyun, Iyer Naveen, and Wang Hongning. 2018. Learning contextual bandits in a non-stationary environment. In Proceeding of the 41st International ACM SIGIR Conference on Research & Development in Information Retrieval. 495–504.
60. [60] Xu Kerui, Yang Jingxuan, Xu Jun, Gao Sheng, Guo Jun, and Wen Ji-Rong. 2021. Adapting user preference to online feedback in multi-round conversational recommendation. In Proceedings of the 14th ACM International Conference on Web Search and Data Mining. 364–372.
61. [61] Yu Tong, Shen Yilin, and Jin Hongxia. 2019. A visual dialog augmented interactive recommender system. In Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 157–165.
62. [62] Yuan Xiaofeng, Han Lixin, Qian Subin, Zhu Licai, Zhu Jun, and Yan Hong. 2021. Preliminary data-based matrix factorization approach for recommendation. Inf. Process. Manage. (2021).
63. [63] Zachariah Dave, Sundin Martin, Jansson Magnus, and Chatterjee Saikat. 2012. Alternating least-squares for low-rank matrix reconstruction. IEEE Sign. Process. Lett. (2012), 231–234.
64. [64] Zamani Hamed, Dumais Susan, Craswell Nick, Bennett Paul, and Lueck Gord. 2020. Generating clarifying questions for information retrieval. In WWW, 418–428.
65. [65] Zhang Xiaoying, Xie Hong, Li Hang, and Lui C.S. John. 2020. Conversational contextual bandit: Algorithm and application. In Proceedings of the Web Conference 2020. 662–672.
66. [66] Zhang Yongfeng, Chen Xu, Ai Qingyao, Yang Liu, and Croft Bruce W.. 2018. Towards conversational search and recommendation: System ask, user respond. In Proceedings of the 27th ACM International Conference on Information and Knowledge Management. 177–186.
67. [67] Zhou Chunyi, Jin Yuanyuan, Wang Xiaoling, and Zhang Yingjie. 2020. Conversational music recommendation based on bandits. In IEEE International Conference on Knowledge Graph (ICKG’20). IEEE, 41–48.
68. [68] Zhou Kun, Zhao Xin, Bian Shuqing, Zhou Yuanhang, Wen Ji-Rong, and Yu Jingsong. 2020. Improving conversational recommender systems via knowledge graph based semantic fusion. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 1006–1014.
69. [69] Zhu Tianyu, Sun Leilei, and Chen Guoqing. 2021. Graph-based embedding smoothing for sequential recommendation. IEEE Trans. Knowl. Data Eng. 35, 1 (2021), 496–508.
70. [70] Zhu Yadong, Wang Xiliang, Li Qing, Yao Tianjun, and Liang Shangsong. 2021. Botspot++: A hierarchical deep ensemble model for bots install fraud detection in mobile advertising. ACM Trans. Inf. Syst. 40, 3 (2021), 1–28.
71. [71] Zou Jie and Kanoulas Evangelos. 2019. Learning to ask: Question-based sequential bayesian product search. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management. 369–378.
72. [72] Zou Jie, Li Dan, and Kanoulas Evangelos. 2018. Technology assisted reviews: Finding the last few relevant documents by asking yes/no questions to reviewers. In Proceeding of the 41st International ACM SIGIR Conference on Research & Development in Information Retrieval. 949–952.
73. [73] Zou Jie, Yifan Chen, and Evangelos Kanoulas. 2020. Towards question-based recommender systems. In Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. 881–890.

Index Terms

• Information systems

  • Information retrieval

    • Retrieval tasks and goals

      • Recommender systems