Collaborative filtering (CF) is a successful approach commonly used by many recommender systems. Conventional CF-based methods use the ratings given to items by users as the sole source of information for learning to make recommendation. However, the ratings are often very sparse in many applications, causing CF-based methods to degrade significantly in their recommendation performance. To address this sparsity problem, auxiliary information such as item content information may be utilized. Collaborative topic regression (CTR) is an appealing recent method taking this approach which tightly couples the two components that learn from two different sources of information. Nevertheless, the latent representation learned by CTR may not be very effective when the auxiliary information is very sparse. To address this problem, we generalize recently advances in deep learning from i.i.d. input to non-i.i.d. (CF-based) input and propose in this paper a hierarchical Bayesian model called collaborative deep learning (CDL), which jointly performs deep representation learning for the content information and collaborative filtering for the ratings (feedback) matrix. Extensive experiments on three real-world datasets from different domains show that CDL can significantly advance the state of the art.
CDL Variants (Direct Extensions of CDL)
- Deep Collaborative Filtering via Marginalized Denoising Auto-encoder
by Li et al., CIKM 2015.
- Collaborative Deep Ranking: A Hybrid Pair-Wise Recommendation Algorithm with Implicit Feedback
by Ying et al., PAKDD 2016.
- Hybrid Recommender System based on Autoencoders
by Strub et al., RecSys DLRS Workshop 2016.
- Collaborative Knowledge Base Embedding for Recommender Systems
by Zhang et al., KDD 2016.
- Ask the GRU: Multi-Task Learning for Deep Text Recommendations
by Bansal et al., RecSys 2016.
- Convolutional Matrix Factorization for Document Context-Aware Recommendation
by Kim et al., RecSys 2016.
- Collaborative Denoising Auto-Encoders for Top-N Recommender Systems
by Wu et al., WSDM 2016.
- Collaborative Recurrent Autoencoder: Recommend while Learning to Fill in the Blanks
by Wang et al., NIPS 2016.
- Joint Deep Modeling of Users and Items Using Reviews for Recommendation
by Zheng et al., WSDM 2017.
- Collaborative Metric Learning
by Hsieh et al., WWW 2017.
- A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems
by Dong et al., AAAI 2017.
- Neural Rating Regression with Abstractive Tips Generation for Recommendation
by Li et al., SIGIR 2017.
- AutoSVD++: An Efficient Hybrid Collaborative Filtering Model via Contractive Auto-encoders
by Zhang et al., SIGIR 2017.
- Collaborative Variational Autoencoder for Recommender Systems
by Li et al., KDD 2017.
- Spatial-Aware Hierarchical Collaborative Deep Learning for POI Recommendation
by Yin et al., TKDE 2017.
- DropoutNet: Addressing Cold Start in Recommender Systems
by Volkovs et al., NIPS 2017.
Earlier Related Works and Other Follow-up Works
- Restricted Boltzmann Machines for Collaborative Filtering
by Salakhutdinov et al., ICML 2007.
- A non-IID Framework for Collaborative Filtering with Restricted Boltzmann Machines
by Georgiev and Nakov, ICML 2013.
- Deep Content-based Music Recommendation
by van den Oord et al., NIPS 2013.
- Improving Content-based and Hybrid Music Recommendation using Deep Learning
by Wang and Wang, ACM MM 2014.
- Deep Coevolutionary Network: Embedding User and Item Features for Recommendation
by Dai et al., RecSys DLRS Workshop 2016.
- Improved Recurrent Neural Networks for Session-based Recommendations
by Tan et al., RecSys DLRS Workshop 2016.
- Wide & Deep Learning for Recommender Systems
by Cheng et al., RecSys DLRS Workshop 2016.
- Triply Stochastic Variational Inference for Non-linear Beta Process Factor Analysis
by Fan et al., ICDM 2016.
- Session-based Recommendations with Recurrent Neural Networks
by Hidasi et al., ICLR 2016.
- Expert Finding for Community-Based Question Answering via Ranking Metric Network Learning
by Zhao et al., IJCAI 2016.
- Parallel Recurrent Neural Network Architectures for Feature-rich Session-based Recommendations
by Hidasi et al., RecSys 2016.
- Deep Neural Networks for YouTube Recommendations
by Covington et al., RecSys 2016.
- Multi-Rate Deep Learning for Temporal Recommendation
by Song et al., SIGIR 2016.
- Neural Collaborative Filtering
by He et al., WWW 2017.
Please do not hesitate to inform us of any missing works.
Below is the recall@M for different methods including our CDL. M is the number recommended items.
Here is the mean average precision (mAP) with the cut-off point set to 500.
Interpretability and Accuracy
Below are two case studies where CDL and CTR recommend articles to User I and User II.
Sensitivity to Users' Interest Change and Accuracy
Below is a third case study where an example user (User III) 'watched' 10 movies one by one. During the process, CDL and CTR recommend movies to the user.
Movie 1 and Movie 2: The first 2 movies User III watched are 'Moonstruck' and 'True Romance', both of which are romance movies. Naturally, both CDL and CTR recommend some related romance movie to the user.
Movie 3 and Movie 4: The next 2 movies User III watched are 'Johnny English' and 'American Beauty'. These are drama and action movies, which are very different from the previous two. Note that this indicates a change of user interest. CDL sensitive enought to capture this change and recommend both romance and drama/action movies to User III, boosting the accuracy from 30% to 50%. However, CTR fails to do so and the accuracy remains 20%.
Other Movies: After User III watched and liked 10 movies, the accuracy of CDL's and CTR's recommendations is 50% and 90%, respectively.
Brief Model Introduction
In CDL, a probabilistic stacked denoising autoencoder (pSDAE) is connected to a regularized probabilistic matrix factorization (PMF) component to form a unified probabilistic graphical model.
Model training will alternate between pSDAE and regularized PMF. In each epoch, a pSDAE with a reconstruction target at the end and a regression target in the bottleneck will be udpated before updating the latent factors U and V in the regularized PMF.
Below is the graphical model for CDL. The part in the red rectangle is pSDAE and the rest is the PMF component regularized by pSDAE. Essentially, the updating will alternate between pSDAE (updating W and b) and the regularized PMF component (updating U and V).
For more details on the notation, please refer to this page or the paper.
Code and Data
The code is available on Github or here, and the data is also available.
You can also check out our MXNet code for a simplied version of CDL on Github or here.
We also made a python notebook as part of the KDD 2016 MXNet tutorial, which is available on Github.
Other implementations (third-party):
Tensorflow code by gtshs2.
Keras code by zoujun123.
Python code by xiaoouzhang.
This research has been partially supported by research grant FSGRF14EG36.