Multi-Task Feature Learning for Knowledge Graph Enhanced Recommendation

By Hongwei Wang, Fuzheng Zhang, Miao Zhao, Wenjie Li, Xing Xie, Minyi Guo. The Web Conference, 2019

536 citations recorded by the source corpus.

Collaborative filtering often suffers from sparsity and cold start problems in real recommendation scenarios, therefore, researchers and engineers usually use side information to address the issues and improve the performance of recommender systems. In this paper, we consider knowledge graphs as the source of side information. We propose MKR, a Multi-task feature learning approach for Knowledge graph enhanced Recommendation. MKR is a deep end-to-end framework that utilizes knowledge graph embedding task to assist recommendation task. The two tasks are associated by crosscompress units, which automatically share latent features and learn high-order interactions between items in recommender systems and entities in the knowledge graph. We prove that crosscompress units have sufficient capability of polynomial approximation, and show that MKR is a generalized framework over several representative methods of recommender systems and multi-task learning. Through extensive experiments on real-world datasets, we demonstrate that MKR achieves substantial gains in movie, book, music, and news recommendation, over state-of-the-art baselines. MKR is also shown to be able to maintain satisfactory performance even if user-item interactions are sparse.

Subfields: Artificial Intelligence, Information Systems

View Paper: https://www.semanticscholar.org/paper/4e810d6b04de4e657a4cda2f58296b01e91eae53

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. Xing Xie
2. Wayne Xin Zhao
3. Philip S. Yu
4. Xiaohan Zou
5. Hongwei Wang
6. Wenjie Li
7. Ji-rong Wen
8. Miao Zhao

In Information Systems

1. Xing Xie
2. Xiaohan Zou
3. Wayne Xin Zhao
4. Wenjie Li
5. Fei Sun
6. Hongwei Wang
7. Bin Cui
8. Shiwen Wu

Citations in the corpus

• A Survey on Accuracy-Oriented Neural Recommendation: From Collaborative Filtering to Information-Rich Recommendation (2021, unknown): “[178], [179], and GNN-based approaches [97], [128], [141], [170], [180], [181], [182], [183], [184], [185], [186], [187].”
• A Survey on Knowledge Graph-Based Recommender Systems (2020, methodology): “[2], [14], [44], [45], [66], [70], [81], [83], [87], [92], [93], [95], [96] MovieLens-20M [44], [86], [88], [89], [91] museums, parks, cities, etc.) to users based on their historical check-in data.”
• A Survey on Knowledge Graphs: Representation, Acquisition, and Applications (2020, background): “To enable end-to-end training, MKR [205] associates multi-task knowledge graph representation and recommendation by sharing latent features and modeling high-order item-entity interaction.”
• RecBole: Towards a Unified, Comprehensive and Efficient Framework for Recommendation Algorithms (2020, methodology): “e. Category Model Reference General recommendation Popularity - ItemKNN Deshpande and Karypis (2004) BPR Rendle et al. (2009) NeuMF He et al. (2017b) DMF Xue et al. (2017) NAIS He et al. (2018b) NGCF Wang et al. (2019e) GCMC van den Berg et al. (2017) LightGCN He et al. (2020) DGCF Wang et al. (2020”
• Knowledge-aware Graph Neural Networks with Label Smoothness Regularization for Recommender Systems (2019, methodology): “classified into three categories: (1) Embedding-based methods [9, 26, 27, 34] pre-process a KG with knowledge graph embedding (KGE) TransE [2] assumes head + relation = tail ), which are more suitable (2) Path-based methods [8, 33, 36] (3) Hybrid methods [18, 24, 28] combine the above two…”
• Graph Neural Networks in Recommender Systems: A Survey (2020, methodology): “Previous works preprocess knowledge graph by knowledge graph embedding (KGE) methods to learn the embeddings of entities and relations, such as [26, 146, 196, 202] The limitation of commonly-used KGE methods is that they focus on modeling rigorous semantic relatedness with the transition constraint,”
• Explainable Interaction-driven User Modeling over Knowledge Graph for Sequential Recommendation (2019, methodology): “The simple KGE method is used to embed entities and relationships into continuous vector space while preserving their structure information [24].”
• A Survey on Application of Knowledge Graph (2020, background): “[26] designs a cross & compress unit to automatically share latent features between KGE task and recommendation task, and learn high-order interactions between items in recommender systems and entities in the KG.”
• Knowledge Graphs: Opportunities and Challenges (2023, methodology): “Recently, entity alignment has been the primary method for implementing knowledge fusion tasks.”
• Hard Negative Sampling via Large Language Models for Recommendation (2025, unknown): “Popular-based methods tend to select more frequent items as negatives (Cheng et al., 2021; Wang et al., 2019a).”