Improving Conversational Recommender Systems via Knowledge Graph based Semantic Fusion

By Kun Zhou, Wayne Xin Zhao, Shuqing Bian, Yuanhang Zhou, Ji-rong Wen, Jingsong Yu. Knowledge Discovery and Data Mining, 2020

404 citations recorded by the source corpus.

Conversational recommender systems (CRS) aim to recommend high-quality items to users through interactive conversations. Although several efforts have been made for CRS, two major issues still remain to be solved. First, the conversation data itself lacks of sufficient contextual information for accurately understanding users' preference. Second, there is a semantic gap between natural language expression and item-level user preference. To address these issues, we incorporate both word-oriented and entity-oriented knowledge graphs~(KG) to enhance the data representations in CRSs, and adopt Mutual Information Maximization to align the word-level and entity-level semantic spaces. Based on the aligned semantic representations, we further develop a KG-enhanced recommender component for making accurate recommendations, and a KG-enhanced dialog component that can generate informative keywords or entities in the response text. Extensive experiments have demonstrated the effectiveness of our approach in yielding better performance on both recommendation and conversation tasks.

Subfields: Artificial Intelligence, Information Systems

View Paper: https://www.semanticscholar.org/paper/2275558bfc6003fd7b146a86a70f3512a33f5874

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

1. Xiangnan He
2. Wayne Xin Zhao
3. Tat-seng Chua
4. Ji-rong Wen
5. Xiang Wang
6. Kun Zhou
7. Fuzheng Zhang
8. Zhijin Qin

In Information Systems

1. Xiangnan He
2. Wayne Xin Zhao
3. Tat-seng Chua
4. Ji-rong Wen
5. Xiang Wang
6. Fuzheng Zhang
7. Kun Zhou
8. Wenqiang Lei

Citations in the corpus

• S3-Rec: Self-Supervised Learning for Sequential Recommendation with Mutual Information Maximization (2020, methodology): “For this problem, we are inspired by the recently proposed mutual information maximization (MIM) method [5, 10, 11, 30].”
• CRSLab: An Open-Source Toolkit for Building Conversational Recommender System (2021, methodology): “To fill these gaps, based on previous works (Lei et al., 2020; Zhou et al., 2020c; Sun and Zhang, 2018), we unify the task of CRS into two basic sub-tasks and an auxiliary sub-task, namely recommendation, conversation and policy.”
• Advances and Challenges in Conversational Recommender Systems: A Survey (2021, background): “250 These efforts can also be found in the field of proactive conversation [140, 227, 6].”
• Beyond Transmitting Bits: Context, Semantics, and Task-Oriented Communications (2022, positive): “The knowledge graph based semantics has also been employed in data integration [31], recommendation systems [32], and realtime ranking [33]–[35].”
• How Can Recommender Systems Benefit from Large Language Models: A Survey (2023, neutral): “However, they only leverage small-scale language models as feature encoders, and thus the key capacities ( e.g. , reasoning, instruction following) of large foundation models still remain underexplored in this quadrant.”