SECRS:基于语义增强的对话推荐系统

doi:10.19678/j.issn.1000-3428.0252002

摘要/Abstract

摘要： 对话推荐系统(CRS)根据用户的多轮交互提供个性化的推荐,在电子商务和电影咨询等领域具有广阔的应用前景。然而现有工作未针对性地提取实际对话中项目间的关联信息,同时在语义融合时未考虑单词与实体共存情况的权重分配,导致系统不能全面提取对话文本所反映的语义信息,进而影响系统的对话生成能力与个性化推荐能力。因此,本文提出一种基于语义增强的对话推荐系统(SECRS),构建可调优的提示来指示预训练语言模型DialoGPT完成对话推荐任务。首先提出会话增强的项目嵌入提取方法,利用图神经网络挖掘实际对话中项目间的转承和共现关系来增强项目的语义。其次提出自适应的语义融合方法,将会话增强的项目嵌入与实体嵌入进行加权相加,再通过独立映射与联合融合的方式将会话增强前后的实体嵌入分别与单词嵌入进行融合,构成对话任务和推荐任务中语义增强的提示,从而提升系统的语言表达能力和个性化推荐能力。在两个数据集上的实验结果表明,SECRS优于多数主流的对话推荐系统,与基线UniCRS相比,SECRS在ReDIAL数据集上的Recall@10指标提高8.33%,Distinct-4指标提高45.15%。对比实验与消融实验的结果表明,SECRS能生成更精确的推荐结果及语言更丰富的回复。

Abstract: Conversational Recommender System (CRS) can provide personalized recommendations through multi-round user interactions. It has broad application prospects in fields such as e-commerce and movie consultation. However, existing studies neither specifically extract the association information between items in actual dialogues nor consider the weight distribution when words and entities coexist. This results in the inability of the system to comprehensively extract the semantic information reflected in the conversation text and consequently affects its ability to generate fluent responses and personalized recommendations. To address these limitations, this study proposes Semantic Enhanced Conversational Recommender System (SECRS), which tunes the pretrained language model DialoGPT with tunable prompts to perform conversational recommendation tasks. First, this study develops a session-enhanced item extraction method using graph neural networks to enhance the semantic information of item-level entities, which mines the transitivity and co-occurrence relations among items. Second, it introduces an adaptive semantic fusion method, which weighs session-enhanced item embeddings and entity embeddings, then fuses the entity embeddings before and after session enhancement with the word embeddings by independent mapping and joint fusion, respectively, to constitute the semantic-enhanced prompts in the conversational task and recommendation task. These methods improve the linguistic expressions and personalized recommendation capabilities of the system. The experimental results on two datasets show that SECRS outperforms the mainstream CRSs. Compared with the baseline model UniCRS, SECRS achieves an 8.33% increase in Recall@10 and a 45.15% improvement in Distinct-4 on the ReDIAL dataset. Comparison and ablation experiments demonstrate that the SECRS can generate responses with more satisfying recommendations and a more diverse language.

Key words: conversational recommendation, semantic fusion, prompt learning, pretrained language model, graph neural networks

中图分类号:

TP391.3

刘春, 王汝婷, 林泓. SECRS:基于语义增强的对话推荐系统[J]. 计算机工程, 2026, 52(10): 94-105.

LIU Chun, WANG Ruting, LIN Hong. SECRS: Semantic Enhanced Conversational Recommender System[J]. Computer Engineering, 2026, 52(10): 94-105.

https://www.ecice06.com/CN/Y2026/V52/I10/94

参考文献

[1] 张宝鑫, 杨丹, 聂铁铮, 等. 基于自监督的多视角图协同过滤推荐方法[J]. 计算机工程, 2024, 50(5): 100-110. ZHANG B X, YANG D, NIE T Z, et al. Recommendation method based on self-supervised multi-view graph collaborative filtering[J]. Computer Engineering, 2024, 50(5): 100-110. (in Chinese)
[2] 杨兴耀, 马帅, 张祖莲, 等. 基于偏好感知的去噪图卷积网络社交推荐[J]. 计算机工程, 2024, 50(10): 154-163. YANG X Y, MA S, ZHANG Z L, et al. Social recommendation based on preference-aware denoising graph convolutional networks[J]. Computer Engineering, 2024, 50(10): 154-163. (in Chinese)
[3] CHRISTAKOPOULOU K, RADLINSKI F, HOFMANN K. Towards conversational recommender systems[C]//Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, USA: ACM Press, 2016: 815-824.
[4] AUER S, BIZER C, KOBILAROV G, et al. DBpedia: a nucleus for a web of open data[C]//Proceedings of International Semantic Web Conference. Berlin, Germany: Springer, 2007: 722-735.
[5] WANG X L, ZHOU K, WEN J R, et al. Towards unified conversational recommender systems via knowledge-enhanced prompt learning[C]//Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery and Data Mining. New York, USA: ACM Press, 2022: 1929-1937.
[6] GU Y X, HAN X, LIU Z Y, et al. PPT: pre-trained prompt tuning for few-shot learning[C]//Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers). Dublin, Ireland: Association for Computational Linguistics, 2022: 8410-8423.
[7] ZHANG Y Z, SUN S Q, GALLEY M, et al. DIALOGPT: large-scale generative pre-training for conversational response generation[C]//Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics: System Demonstrations. Online: Association for Computational Linguistics, 2020: 270-278.
[8] SCHLICHTKRULL M, KIPF T N, BLOEM P, et al. Modeling relational data with graph convolutional networks[C]//Proceedings of International Semantic Web Conference. Berlin, Germany: Springer, 2018: 593-607.
[9] YANG Y C, CHEN C T, LU T Y, et al. Hierarchical reinforcement learning for conversational recommendation with knowledge graph reasoning and heterogeneous questions[J]. IEEE Transactions on Services Computing, 2023, 16(5): 3439-3452.
[10] 朱立玺, 黄晓雯, 赵梦媛, 等. 基于负反馈修正的多轮对话推荐系统[J]. 计算机学报, 2023, 46(5): 1086-1102. ZHU L X, HUANG X W, ZHAO M Y, et al. Multi-round conversational recommendation system based on negative feedback correction[J]. Chinese Journal of Computers, 2023, 46(5): 1086-1102. (in Chinese)
[11] REN X, YIN H, CHEN T, et al. CRSAL: conversational recommender systems with adversarial learning[J]. ACM Transactions on Information Systems, 2020, 38(4): 1-40.
[12] LI R, KAHOU S, SCHULZ H, et al. Towards deep conversational recommendations[EB/OL]. [2024-12-20]. https://arxiv.org/abs/1812.07617.
[13] ZHANG X Y, XIN X, LI D D, et al. Variational reasoning over incomplete knowledge graphs for conversational recommendation[C]//Proceedings of the 16th ACM International Conference on Web Search and Data Mining. New York, USA: ACM Press, 2023: 231-239.
[14] ZOU J, KANOULAS E, REN P J, et al. Improving conversational recommender systems via transformer-based sequential modelling[C]//Proceedings of the 45th International ACM SIGIR Conference on Research and Development in Information Retrieval. New York, USA: ACM Press, 2022: 2319-2324.
[15] WANG L Z, HU H, SHA L, et al. RecInDial: a unified framework for conversational recommendation with pretrained language models[C]//Proceedings of the 2nd Conference of the Asia—Pacific Chapter of the Association for Computational Linguistics and the 12th International Joint Conference on Natural Language Processing (Volume 1: Long Papers). Online: Association for Computational Linguistics, 2022: 489-500.
[16] HE Z K, XIE Z H, JHA R, et al. Large language models as zero-shot conversational recommenders[C]//Proceedings of the 32nd ACM International Conference on Information and Knowledge Management. New York, USA: ACM Press, 2023: 720-730.
[17] CHEN Q B, LIN J Y, ZHANG Y C, et al. Towards knowledge-based recommender dialog system[C]//Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). Hong Kong, China: Association for Computational Linguistics, 2019: 1803-1813.
[18] LU Y, BAO J W, SONG Y, et al. RevCore: review-augmented conversational recommendation[C]//Proceedings of the Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021. Online: Association for Computational Linguistics, 2021: 1161-1173.
[19] LIN D D, WANG J, LI W J. COLA: improving conversational recommender systems by collaborative augmentation[C]//Proceedings of the AAAI Conference on Artificial Intelligence. [S.l.]: AAAI Press, 2023: 4462-4470.
[20] YANG B W, HAN C, LI Y, et al. Improving conversational recommendation systems’ quality with context-aware item meta-information[C]//Proceedings of the Findings of the Association for Computational Linguistics: NAACL 2022. Seattle, USA: Association for Computational Linguistics, 2022: 38-48.
[21] ZHOU K, ZHAO W X, BIAN S Q, et al. Improving conversational recommender systems via knowledge graph based semantic fusion[C]//Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. New York, USA: ACM Press, 2020: 1006-1014.
[22] ZHOU Y H, ZHOU K, ZHAO W X, et al. C²-CRS: coarse-to-fine contrastive learning for conversational recommender system[C]//Proceedings of the 15th ACM International Conference on Web Search and Data Mining. New York, USA: ACM Press, 2022: 1488-1496.
[23] LIU Y H, OTT M, GOYAL N, et al. RoBERTa: a robustly optimized BERT pretraining approach[EB/OL]. [2024-12-20]. https://arxiv.org/abs/1907.11692.
[24] PANG Y T, WU L F, SHEN Q, et al. Heterogeneous global graph neural networks for personalized session-based recommendation[C]//Proceedings of the 15th ACM International Conference on Web Search and Data Mining. New York, USA: ACM Press, 2022: 775-783.
[25] MüLLER R, KORNBLITH S, HINTON G. When does label smoothing help? [EB/OL]. [2024-12-20]. https://arxiv.org/abs/1906.02629.
[26] SHANG C Z, HOU Y P, ZHAO W X, et al. Multi-grained hypergraph interest modeling for conversational recommendation[J]. AI Open, 2023, 4: 154-164.
[27] RAVAUT M, ZHANG H, XU L, et al. Parameter-efficient conversational recommender system as a language processing task[C]//Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 1: Long Papers). St. Julian’s, Malta: Association for Computational Linguistics, 2024: 152-165.
[28] LIU C, WANG B Q, LI Y Q. Dialog generation model based on variational Bayesian knowledge retrieval method[J]. Neurocomputing, 2023, 561: 126878.

选择文件类型/文献管理软件名称

选择包含的内容