A Multi-View Deep Learning Approach for Cross Domain User Modeling in Recommendation Systems

By Ali Mamdouh Elkahky, Yang Song, Xiaodong He. 2015

735 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Vision and Pattern Recognition, Information Systems, 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 Artificial Intelligence

In Computer Vision and Pattern Recognition

In Information Systems

In Recommender Systems

Citations in the corpus

Neural Collaborative Filtering (2017, methodology): “2) Even though tanh is a better choice and has been widely adopted [6, 44], it only alleviates the issues of sigmoid to a certain extent, since it can be seen as a rescaled version of sigmoid (tanh( x/ 2) = 2 σ ( x ) − 1).”
Deep Neural Networks for YouTube Recommendations (2016, methodology): “used deep learning for cross domain user modeling [5].”
AutoEmb: Automated Embedding Dimensionality Search in Streaming Recommendations (2020, background): “Ali et al.[6] designed Multi-View Deep Neural Network (MV-DNN) which could model the interactions among users and items with multiple domains.”
From Eliza to XiaoIce: challenges and opportunities with social chatbots (2018, background): “…et al., 2014; van den Oord et al., 2016), computer vision (Krizhevsky et al., 2012; He et al., 2016), information retrieval (Huang et al., 2013; Elkahky et al., 2015), multimodal intelligence (Fang et al., 2015; Vinyals et al., 2015; Karpathy and Fei-fei, 2015; He and Deng, 2017), and empathic…”
Joint Deep Modeling of Users and Items Using Reviews for Recommendation (2017, background): “In [9], a multi-view deep model is built to learn the user and item latent factors in a joint manner and map them to a common space.”
xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems (2018, methodology): “[8, 23] employ neural networks to jointly train multiple domains’ latent factors.”
Are we really making much progress? A worrying analysis of recent neural recommendation approaches (2019, neutral): “Non-reproducible: KDD: [43], RecSys: [41], [6], [38], [44], [21], [45], SIGIR: [32], [7], WWW: [42], [11]”
Auto-Encoding User Ratings via Knowledge Graphs in Recommendation Scenarios (2017, background): “Mapping user and items to a latent space, as done in [6] seems to be a good approach to address the recommendation quality in content-based recommender systems.”
Understanding Training Efficiency of Deep Learning Recommendation Models at Scale (2020, background): “Microsoft leverages deep learning recommendation models for news suggestions [16], Netflix for personalized entertainment [47], Alibaba for product recommendations [59].”
Collaborative Knowledge Base Embedding for Recommender Systems (2016, methodology): “For example, Elkahky et al. [8] proposes a multi-view deep learning model to learn user features and item features from different domains, and uses the enriched feature representation to improve the recommendation quality across all the domains.”
Neural Networks and Deep Learning (2018, positive): “A multiview deep learning approach, which has also been extended to temporal recommender systems in a later work [465], is proposed in [110].”
Computing Recommendations via a Knowledge Graph-aware Autoencoder (2018, background): “Analysis about the impact of deep learning on both recommendation quality and system scalability are presented in [13], where the authors first represent users and items through a rich feature set made on different domains and then map them to a latent space.”
Transfer Meets Hybrid: A Synthetic Approach for Cross-Domain Collaborative Filtering with Text (2019, background): “Multiple source domains [32] and multi-view learning [13] are also proposed for integrating information from several domains.”
Joint Representation Learning for Top-N Recommendation with Heterogeneous Information Sources (2017, methodology): “On considering information sources beyond ratings, [2, 36, 41, 45] adopted deep textual modeling on reviews for recommendation, [35, 37] leveraged deep audio embeddings for music recommendation, [40] incorporated knowledge base embedding for recommendation, and [15] studied cross-domain recommendati”
Bridging Collaborative Filtering and Semi-Supervised Learning: A Neural Approach for POI Recommendation (2017, methodology): “For implicit feedback, neural networks have been explored to model context information, such as textual description of items [40], acoustic features of musics [41], cross-domain behaviors of users [9] and the information in knowledge bases [51].”
Cross-domain Recommendation via Deep Domain Adaptation (2018, background): “[8], which dealt with user modeling for cross-domain recommendations using deep neural networks.”