標籤:

Deep Learning in Recommender System

深度學習技術在計算機視覺、語音、自然語言處理方面已經相對成熟,工業界已經開始關注深度學習在推薦系統領域的應用。近三年KDD/WSDM/RecSys等會議出現越來越多的相關論文,這裡備忘深度學習&推薦系統相關的論文列表,供同行追蹤查閱。

(註:這個列表會一直更新,歡迎在評論中補充~)

2014年以前

  1. Salakhutdinov, Restricted Boltzmann Machines for Collaborative Filtering, ICML 2007 (Netflix Prize,SVD++,模型遇到瓶頸,用RBM改進協同過濾)
  2. Recommending Music on Spotify with Deep Learning(Spotify音樂推薦)
  3. Recurrent Neural Networks for Collaborative Filtering (Spotify音樂推薦,CF+DL)
  4. Oord et al., Deep Content-based Music Recommendation, NIPS 2013(用DL提取內容特徵,改進音樂推薦)
  5. Xinxi Wang et al., Improving Content-based and Hybrid Music Recommendation Using Deep Learning, ACM MM 2014 (用DL提取內容特徵,改進音樂推薦)
  6. Session-based Recommendations with Recurrent Neural Networks(Netflix研究員)
  7. A Multi-View Deep Learning Approach for Cross Domain User Modeling in Recommendation Systems, WWW 2015(微軟,新聞推薦領域使用DL的心得)
  8. A Deep Embedding Model for Co-occurrence Learning
  9. The Recommender Problem Revisited, KDD 2014 (推薦系統資深專家Netflix XavierAmatriain的分享,後來在RecSys多次分享Lessons Learned from Building Real--Life Recommender Systems)

2015

  1. Sedhain et al., Autorec: Autoencoders Meet Collaborative Filtering, WWW 2015(用DL無監督Autoencoder改進Item-based CF,預測user對item的評分)
  2. Hao Wang & Naiyan Wang, Collaborative Deep Learning for Recommender Systems, KDD 2015(用DL改進協同過濾)
  3. A Deep Matrix Factorization Method for Learning Attribute Representations

2016

  1. Yao Wu et al., Collaborative Denoising Autoencoders for Top-n Recommender Systems, WSDM 2016
  2. Yin Zheng et al., A Neural Autoregressive Approach to Collaborative Filtering, ICML 2016(Hulu團隊提出用於CF的神經自回歸架構,受啟發於基於CF的RBM和NADE,神經自回歸分布估計器,在Netflix數據集上取得SOTA結果)
  3. Yin Zheng et al., A Deep and Autoregressive Approach for Topic Modeling of Multimodal Data, IEEE Trans on PAMI 2016 (同樣來自Hulu團隊的工作)
  4. Junxuan Chen et al., Deep CTR Prediction in Display Advertising, ACM MM 2016(一個同時考慮圖像特徵和顯示廣告的特徵的深度學習網路)
  5. Covington et al., Deep Neural Networks for YouTube Recommendation, Recsys 2016(Google Youtube視頻推薦系統,two-stage的深度學習框架)
  6. Kim et al., Convolutional Matrix Factorization for Document Context-Aware Recommendation, Recsys 2016 (基於PMF和CNN做文檔推薦)
  7. Wide & Deep Learning for Recommender Systems, 2016(Google PlayApp推薦系統,著名的Wide & Deep模型,融合LR+DNN)
  8. Personal Recommendation Using Deep Recurrent Neural Networks in NetEase, 2016(網易考拉基於RNN和FNN的商品推薦)
  9. Deep Learning over Multi-field Categorical Data - A Case Study on User Response Prediction, ECIR 2016(提出FNN:預訓練FM得到組合特徵,再使用DNN做CTR預估)
  10. Product-based Neural Networks for User Response Prediction, CoRR 2016(提出PNN,在embedding層和全連接層中引入一層product層來學習組合特徵)

2017

  1. Lei Zheng et al., Joint Deep Modeling of Users and Items Using Reviews forRecommendation, WSDM 2017.
  2. Chao-Yuan Wu & Alex Smola, Recurrent Recommender Networks, WSDM 2017
  3. A Hybrid Collaborative Filtering Model with Deep Structure for Recommender Systems, AAAI 2017(攜程提出混合協同過濾和深度AE結構)
  4. DeepFM: A Factorization-Machine based Neural Network for CTR Prediction, IJCAI 2017(華為諾亞方舟,受D&W啟發,提出FM+DNN融合模型,用於手機應用市場的CTR預估)
  5. Deep Matrix Factorization Models for Recommender Systems, IJCAI 2017(南京大學,用DNN改進傳統MF模型,2015有篇DeepMF學習屬性表達的文章)

推薦閱讀:

論文格式排版你真的做對了嗎? 常用格式及其LaTeX書寫方法介紹
如何閱讀論文?——「三十天讀論文計劃」踐行心得

TAG:论文 |