A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems

By Xin Dong, Lei Yu, Zhonghuo Wu, Yuxia Sun, Lingfeng Yuan, Fangxi Zhang. 2017

297 citations recorded by the source corpus.

Collaborative filtering (CF) is a widely used approach in recommender systems to solve many real-world problems. Traditional CF-based methods employ the user-item matrix which encodes the individual preferences of users for items for learning to make recommendation. In real applications, the rating matrix is usually very sparse, causing CF-based methods to degrade significantly in recommendation performance. In this case, some improved CF methods utilize the increasing amount of side information to address the data sparsity problem as well as the cold start problem. However, the learned latent factors may not be effective due to the sparse nature of the user-item matrix and the side information. To address this problem, we utilize advances of learning effective representations in deep learning, and propose a hybrid model which jointly performs deep users and items’ latent factors learning from side information and collaborative filtering from the rating matrix. Extensive experimental results on three real-world datasets show that our hybrid model outperforms other methods in effectively utilizing side information and achieves performance improvement.

Subfields: Information Systems, Representation & Self-Supervised Learning, Recommender Systems

How researchers see this paper

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

In Information Systems

In Representation & Self-Supervised Learning

In Recommender Systems

Citations in the corpus

DKN: Deep Knowledge-Aware Network for News Recommendation (2018, methodology): “to the aforementioned DSSM [18], DeepWide [7], DeepFM [15], YouTubeNet [9] and DMF [50], other popular deep-learning-based recommender systems include Collaborative Deep Learning [45], Hybrid CF-SDAE [12], Multi-view Deep Learning [13], and Neural Collaborative Filtering [16]. The major difference b”
SHINE: Signed Heterogeneous Information Network Embedding for Sentiment Link Prediction (2017, background): “ell as their heterogeneous embedding ei and ej, the predicted sentiment s¯ij can be calculated as s¯ij = f (i,j), where f (·,·)is specific similarity measurement function. For example: •Inner product [3, 5], i.e., s¯ij = eT i j + b, where is a trainable bias parameter; •Euclidean distance [26], i.e.”
Knowledge-aware Autoencoders for Explainable Recommender Systems (2018, methodology): “A similar approach has been proposed in [6], in which they exploit side information in a CF setting by using Stacked Autoencoders in order to overcome the cold start problem and data sparsity.”
Auto-Encoding User Ratings via Knowledge Graphs in Recommendation Scenarios (2017, methodology): “A Hybrid Recommender System is presented in [5].”
Deep Learning Based Recommender System (2017, unknown): “on models. Table 1. A lookup table for reviewed publications. Categories Publications MLP [2, 13, 20, 27, 38, 47, 53, 54, 66, 92, 95, 157, 166, 185], [12, 39, 93, 112, 134, 154, 182, 183] Autoencoder [34, 88, 89, 114, 116, 125, 136, 137, 140, 159, 177, 187, 207], [4, 10, 32, 94, 150, 151, 158, 170, ”
xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems (2018, background): “ing Explicit and Implicit Feature Interactions for Recommender Systems KDD ’18, August 19–23, 2018, London, United Kingdom CF and side information with the purpose of generating better latent factors [7, 39, 45]. [8, 23] employ neural networks to jointly train multiple domains’ latent factors. [3] p”
Computing Recommendations via a Knowledge Graph-aware Autoencoder (2018, background): “Another hybrid approach is proposed in [12]: the aim is to address the sparsity problem using deep learning techniques to model both users and items exploiting side information.”
When large language models meet personalization: perspectives of challenges and opportunities (2023, unknown): “Dong et al. [88] propose a stacked denoising autoencoder that reconstructs item/user ratings and textual information simultaneously, allowing for the joint modeling of collaborative and textual knowledge.”
A Survey of Multi-View Representation Learning (2016, unknown): “[204] propose to jointly perform deep user’s and item’s latent representation learning from side information and collaborative”