Workshop on Graph Neural Networks for Recommendation and Search (GReS) - Naver Labs Europe
GreS workshop

GReS – Workshop on Graph Neural Networks for Recommendation and Search

Co-located with the ACM RecSys ’21 conference. The workshop will be held virtually on October 2nd, 2021.

Paper submission deadline: July 29th, 2021 (AoE)

Follow us on Twitter @GReS_Workshop to stay updated.

GReS Workshop

The longstanding paradigm of collaborative filtering in recommender systems posits that users with similar behavior tend to exhibit similar preferences. A graph formulation naturally arises from this view: the user-item interactions form a bipartite graph, which can be leveraged to refine recommendations by integrating similarities in users’ historical preferences. This perspective inspired numerous graph-based recommendation approaches in the past

Recently, the success brought about by deep learning led to the development of graph neural networks (GNNs). The key idea of GNNs is to propagate high-order information in the graph so as to learn representations which are similar for a node and its neighborhood. GNNs were initially applied to traditional machine learning problems such as classification or regression and later to recommendation and search. GNNs have in particular led to a new state of the art in top-k recommendation and next-item recommendation.

The GReS workshop on Graph Neural Networks for Recommendation and Search is then an endeavor to bridge the gap between the RecSys and GNN communities and promote inter-collaborations, creating a more attractive and dedicated space to foster GNN contributions to the RecSys domain.

The GReS workshop accepts papers of up to 14 pages following the standard single-column ACM RecSys template. This length does not include references and reviewers will be asked to comment on whether the length is appropriate for the contribution. Submissions are double-blind (and therefore should be anonymized) and to be provided in a single file in .pdf format.

Submission link:

Paper submission deadline: July 29th, 2021 (AoE)

Author notification: August 21st, 2021 (AoE)

Camera-ready version deadline: September 4th, 2021 (AoE)


The topics relevant to the GReS workshop include (but are not limited to):

  • Top-k recommendation and matrix completion approaches based on GNNs;
  • Session-based and next-item recommendation via GNNs and dynamic graphs;
  • Knowledge graph and social network-enhanced recommendation models;
  • Multimodal recommendation and search approaches based on GNNs;
  • Explainability, fairness, accountability, transparency, and privacy issues in GNN-based recommendation;
  • GNN-based result diversification in recommendation or search;
  • Hypergraph neural networks for recommendation or search;
  • GNNs for personalized recommendation via link prediction in multipartite or heterogeneous graphs;
  • Temporal and Dynamic GNNs or applications of GNNs to next-item recommendation and dynamic environments;
  • Graph topology inference for recommendation and search;
  • Challenges, pitfalls, and negative results in applying GNNs to recommendation or search;
  • Libraries, benchmarks, and datasets for GNN-based recommendation or search;
  • Industrial applications and scalability of GNNs for recommendation or search.
  • Thibaut Thonet (NAVER LABS Europe)
    Thibaut Thonet is a research scientist at NAVER LABS Europe in the Search & Recommendation group since 2019. Prior to that, he held a postdoctoral position at the University Grenoble Alpes from 2018 to 2019. He obtained his Ph.D. degree in 2017 from the University of Toulouse. He was also visiting researcher at the University of Glasgow in 2014. His academic interests include recommender systems, information retrieval, and machine learning. His research was published at SIGIR, CIKM, and ECIR, and he served as program committee member for conferences such as SIGIR, ECIR, WebConf, AAAI, and NeurIPS. He received an outstanding reviewer award at SIGIR ’17 and a recognized reviewer award at ECIR ’19.
  • Stéphane Clinchant (NAVER LABS Europe)
    Stéphane Clinchant is a senior research scientist at NAVER LABS Europe. He obtained his PhD in 2011 in Information Retrieval from the University of Grenoble. He has worked on various European research projects, technology transfer to business groups to deploy machine learning algorithms and challenges such as CLEF, VisDA and TREC. He recently visited the NAVER Papago team in South Korea to work on Neural Machine Translation. He started to work on GNNs in 2017 on the topic of cross modal retrieval and document layout analysis. His research has been published at SIGIR, CIKM, KDD, ECIR, ACL and his interests include information retrieval, machine translation, machine learning and graph neural networks.
  • Carlos Lassance (NAVER LABS Europe)
    Carlos Lassance joined NAVER LABS Europe as a Research Scientist in 2020. He received an engineering double-degree from PUC-Rio and IMT Atlantique in 2017, and then completed his Ph.D. in machine learning and graphs from IMT Atlantique in 2020. During his PhD he developed various methods to improve machine learning tasks using graph-based models. He recently joined NAVER LABS Europe as a research scientist to develop graph-based methods for recommendation systems. His main research interests are Graph Signal Processing (GSP), information retrieval and deep neural networks.
  • Elvin Isufi (Delft University of Technology)
    Elvin Isufi is an assistant professor in the Intelligent System department of the Delft University of Technology (TU Delft), where he co-founded and co-directs AIdroLab –one of the TU Delft AI labs focusing on graph learning research for water networks. His main research interest are on fundamental aspects about processing and learning over networks and he is particularly focusing on applications in recommender systems. Elvin received the Ph.D. in 2019 from TU Delft. He has been a postdoctoral researcher at the University of Pennsylvania and visiting researcher at EPFL and University of Perugia. Elvin is currently the Lead Guest Editor of the special Issue entitled “Processing and Learning over Graphs” in Elsevier Signal Processing and was the organizer of the special issue with the homonym title in EURASIP EUSIPCO 2020 –largest European signal processing conference. He has given tutorial about graph signal processing and graph neural networks in ICASSP 2020 –the flagship signal processing conference according to IEEE Signal Processing Society– and SPCOM 2020. He was part of the organizing committee of EURASIP EUSIPCO 2020 and is also part of the organizing committee of the same conference proposal for 2024.
  • Jiaqi Ma (University of Michigan)
    Jiaqi Ma is a Ph.D. candidate at University of Michigan. He has done work in the areas of graph representation learning and neural recommender system in his PhD study and his internships at Google Brain. His research has been published in major AI conferences, including ICLR, NeurIPS, KDD, WWW, AISTATS, etc. He co-organized the workshop on Graph Learning Benchmark at the Web Conference 2021, and regularly served as reviewers in AI-related journals and conferences.
  • Yutong Xie (University of Michigan)
    Yutong Xie is a Ph.D. student at University of Michigan. Prior to this, she received her Bachelor’s degree at Shanghai Jiao Tong University. Her research interests generally lie in machine learning methods for both explicitly and implicitly structured data, especially for graphs and texts. She has also researched in the area of AI powered drug discovery during her internship at ByteDance AI Lab.
  • Jean-Michel Renders (NAVER LABS Europe)
    Jean-Michel Renders is Principal Scientist at NAVER LABS Europe since 2017. He leads the “Search and Recommendation’” team  and most of his research topics are related to machine learning applied to the exploitation of unstructured data (text, speech, images, etc.) from multiple, heterogeneous sources (social media, knowledge bases, transactional databases, speech transcripts or signals from business conversations, etc.). Prior to that, he was Principal Scientist at Xerox Research Centre Europe. He holds a PhD from the University of Brussels on the theme of Deep Learning applied to Process Control.
  • Michael Bronstein (Imperial College London / Twitter)
    Michael Bronstein (PhD Technion, 2017) is a professor at Imperial College London, where he holds the Chair in Machine Learning and Pattern Recognition, and Head of Graph Learning Research at Twitter. His main research interests are in geometric ML and graph representation learning. Michael is a Member of the Academia Europaea, Fellow of IEEE, IAPR, and ELLIS, ACM Distinguished Speaker, World Economic Forum Young Scientist, and a co-Director of the ELLIS Program on Geometric Deep Learning. He is one of the pioneers of graph learning methods and has spearheaded their dissemination and popularization by organizing multiple workshops and tutorials on these topics. In addition to his academic career, Michael is a serial entrepreneur and founder of multiple startup companies, including Fabula AI (acquired by Twitter in 2019) that pioneered the use of GNNs for misinformation detection.

3:00 – 3:05 PM: Workshop’s Opening
3:05 – 3:45 PM: Keynote I – Graph Neural Networks for Recommendation by Xiang Wang (National University of Singapore)
3:45 – 3:50 PM: Short Break
3:50 – 4:10 PM: Accepted Paper I – Knowledge Graph Attention for Sequential Recommendations by Amjadi et al
4:10 – 4:50 PM: Keynote II – Towards GNN explainability: Random Walk Graph Neural Networks by Michalis Vazirgiannis (Ecole Polytechnique)
4:50 – 5:20 PM: Long Break on
5:20 – 5:40 PM: Accepted Paper II – CoRGi: Content-Rich Graph Neural Networks with Attention by Kim et al
5:40 – 6:00 PM: Accepted Paper III – Referral prediction in Healthcare using Graph Neural Networks by Duarte et al
6:00 – 6:10 PM: Short Break
6:10 – 6:50 PM: Keynote III – Powering Pinterest Recommendations with Graph Neural Networks by Andrew Zhai (Pinterest)
6:50 – 7:00 PM: Workshop’s Closing

Xiang Wang (National University of Singapore)
Title: Graph Neural Networks for Recommendation
Abstract: Graph Neural Networks (GNNs) have achieved remarkable success in many domains and shown great potentials in personalized recommendation. In this talk, I will give a brief introduction on why GNNs are suitable for recommendation, and how to incorporate GNNs into some recommendation scenarios (e.g., collaborative filtering, knowledge-aware recommendation). I will also share some points on what advances we are researching into (e.g., self-supervised learning, contrastive learning). In the last part, I will discuss in the “Achilles’ Heel” of the GNN-based recommender models (e.g., amplify popularity bias), as well as some possible solutions (e.g., causal inference).


Michalis Vazirgiannis (Ecole Polytechnique)
Title: Towards GNN explainability: Random Walk Graph Neural Networks
Abstract: In recent years, graph neural networks (GNNs) have become the de facto tool for performing machine learning tasks on graphs. Most GNNs belong to the family of message passing neural networks (MPNNs). These models employ an iterative neighborhood aggregation scheme to update vertex representations. Then, to compute vector representations of graphs, they aggregate the representations of the vertices using some permutation invariant function. One would expect the hidden layers of a GNN to be composed of parameters that take the form of graphs. However, this is not the case for MPNNs since their update procedure is parameterized by fully-connected layers. In this paper, we propose a more intuitive and transparent architecture for graph-structured data, so-called Random Walk Graph Neural Network (RWNN). The first layer of the model consists of a number of trainable “hidden graphs” which are compared against the input graphs using a random walk kernel to produce graph representations. These representations are then passed on to a fully-connected neural network which produces the output. The employed random walk kernel is differentiable, and therefore, the proposed model is end-to-end trainable. We demonstrate the model’s transparency on synthetic datasets. Furthermore, we empirically evaluate the model on graph classification datasets and show that it achieves competitive performance.


Andrew Zhai (Pinterest)
Title: Powering Pinterest Recommendations with Graph Neural Networks
Abstract: Pinterest is the home of inspiration to over 450M monthly active users. Inspiration comes from our personalized recommender systems generating content from our catalog of billions of ideas. Graph Neural Networks (GNNs) have been a key method in improving the predictive performance of these systems through helping us understand content, search queries, and users more comprehensively. In our talk we will discuss the evolution of these representations starting with (1) using GNNs to model content (pins) combining multimodal node features with our web scale pin-board graph of billions of nodes and edges (2) showing how our content embeddings enable us to learn good representations of search queries and users, combining our GNN embeddings with techniques such as sequence models to capture temporal behavior of users across all of Pinterest. Beyond sharing technical details, we will quantify our learnings with online experimentation showing the impact of our methods.

This web site uses cookies for the site search, to display videos and for aggregate site analytics.

Learn more about these cookies in our privacy notice.


Cookie settings

You may choose which kind of cookies you allow when visiting this website. Click on "Save cookie settings" to apply your choice.

FunctionalThis website uses functional cookies which are required for the search function to work and to apply for jobs and internships.

AnalyticalOur website uses analytical cookies to make it possible to analyse our website and optimize its usability.

Social mediaOur website places social media cookies to show YouTube and Vimeo videos. Cookies placed by these sites may track your personal data.