Explore

RobustRecSys @ RecSys2024

Workshop on Design, Evaluation and Deployment of Robust Recommender Systems

More information to come: stay tuned!

Find the important dates

here⁠

⁠

Find the call for papers

here⁠

⁠

Find the schedule

here⁠

⁠

Overview

In recent years, recommender systems have become indispensable tools in various domains, aiding users in discovering relevant content amidst the overwhelming amount of available material.

However, the effectiveness and reliability of these systems are often hindered by various challenges such as data perturbations, missing data, noise, and bias. In this workshop, we aim to explore and address these challenges by focusing on the development of robust recommender systems.

Robustness in recommender systems refers to their ability to maintain performance and effectiveness under adverse conditions, including unexpected variations in the data environment.

By fostering discussions and collaborations among researchers and practitioners, this workshop seeks to advance the state-of-the-art in robust recommender systems, thereby enhancing their usability and trustworthiness in real-world applications.

Important Dates

Submission deadline: August 31st, 2024

Acceptance Notification: September 13th, 2024

RobustRecSys Workshop: October 18th, 2024

Camera-ready versions of accepted papers due: TBD

Deadlines refer to 23:59 (11:59pm) in the AoE (Anywhere on Earth) time zone.

Topics

The workshop centers on the theme of enhancing the robustness of recommender systems, a multifaceted challenge encompassing resilience to missing data, noise, and bias. Robustness in recommender systems can be assessed not only by their performance metrics but also by their stability in the face of various perturbations, with recent studies highlighting the vulnerability of widely used recommender systems to perturbations in training data.

Our workshop aims to address these challenges by focusing on three key areas:

Identifying and quantifying perturbations present in the data used to train recommender systems

Understanding the impact of these perturbations on system performance, assessed using both standard metrics and specialized robustness metrics

Exploring strategies to mitigate the effects of these perturbations and enhance the robustness of recommender systems against them

By concentrating on these aspects, we aim to provide attendees with practical insights and methodologies to improve the reliability and resilience of recommender systems in real-world applications.

In particular, we are looking for contributions that study:

Robustness against Training Data Perturbations: Recommender systems often rely on large-scale datasets for training. However, these datasets may be subject to noise, outliers, and adversarial attacks, leading to suboptimal performance. This workshop will explore techniques for enhancing the robustness of recommender systems against such perturbations, including robust optimization, data augmentation, and outlier detection methods.

Handling Missing Data: In real-world scenarios, recommender systems frequently encounter missing or incomplete data, which can degrade their performance. Strategies for effectively handling missing data, such as imputation methods, probabilistic modeling, and collaborative filtering techniques, will be discussed to improve the robustness of recommender systems.

Mitigating Noise and Bias: Noise and bias in data can introduce inaccuracies and unfairness in recommendations, affecting user satisfaction and trust. This workshop will investigate approaches for mitigating noise and bias in recommender systems, including fairness-aware recommendation algorithms, debiasing techniques, and post-processing methods.

Evaluating Robustness: Assessing the robustness of recommender systems is crucial for understanding their performance under diverse conditions. This workshop will explore evaluation metrics and methodologies tailored for measuring the robustness of recommender systems, considering factors such as adversarial robustness, generalization capabilities, and resilience to concept drift.

Applications and Case Studies: Real-world applications of robust recommender systems across various domains, including e-commerce, social media, healthcare, and education, will be presented and discussed. Case studies highlighting the practical challenges and solutions in deploying robust recommender systems will provide valuable insights for both researchers and practitioners.

Other: One of the goals of this workshop is to collect new ideas and challenges, so proposals in this sense are very much welcomed.

Call for papers

We are pleased to invite you to contribute to the First Workshop on Design, Evaluation and Deployment of Robust Recommender Systems @ RecSys 2024, the premier venue for research on the foundations and applications of recommendation technologies. Each accepted paper is expected to be presented in person. All accepted papers will be published in

CEUR-WS.org⁠

proceedings.

We invite submissions of original research on the main themes of the workshop, identified above. We encourage different types of papers, from theoretical ones, focusing on the mathematics behind Recommender Systems, to industrial papers that focus on open challenges in their specific environment.

Reviewing Process

All submissions will be peer-reviewed (double-blind) by the program committee and judged by their relevance to the workshop, especially to the main themes identified above, and their potential to generate discussion.

Submission Guidelines

All submissions must be written in English and formatted according to the latest ACM’s archival publication format. All authors should submit manuscripts for review in a single-column format. Instructions for Word and LaTeX authors are given below:

Microsoft Word: Write your paper using the

Submission Template⁠

(Review Submission Format). Follow the embedded instructions to apply the paragraph styles to your various text elements. The text is in single-column format at this stage and no additional formatting is required.

LaTeX: Please use the latest version of the

Primary Article Template⁠

– LaTeX to create your submission. You must use the “manuscript” option with the \documentclass[manuscript,anonymous]{acmart} command to generate the output in a single-column format which is required for review. Please see the

LaTeX documentation⁠

and

ACM’s LaTeX best practices guide⁠

for further instructions. To ensure 100% compatibility with The ACM Publishing System (TAPS), please restrict the use of packages to the

whitelist of approved LaTeX packages⁠

Submissions must describe work that is not previously published, not accepted for publication elsewhere, and not currently under review elsewhere.

Please note that at least one of the authors of each accepted paper must register for the workshop and present the paper.

We invite research contributions, position, demo and opinion papers. Submissions must either be short (at most 4 pages, including appendices) or full papers (at most 8 pages, including appendices). References do not count against the page limit.

We encourage but do not require authors to release any code and/or datasets associated with their paper.

All submissions and reviews will be handled electronically. Papers must be submitted via

EasyChair⁠

https://easychair.org/my/conference?conf=recsys2024workshops⁠

⁠

Accepted papers

Addressing bias in Recommender Systems: A Case Study on Data Debiasing Techniques in Mobile Games

Yixiong Wang, Maria Paskevich and Hui Wang

Robust Training Objectives Improve Embedding-based Retrieval in Industrial Recommendation Systems

Matthew Kolodner, Mingxuan Ju, Zihao Fan, Tong Zhao, Elham Ghazizadeh, Yan Wu, Neil Shah and Yozen Liu

Reducing Popularity Influence by Addressing Position Bias

Andrii Dzhoha, Alexey Kurennoy, Vladimir Vlasov and Marjan Celikik

The Role of Fake Users in Sequential Recommender Systems

Filippo Betello

Removing Bad Influence: Identifying and Pruning Detrimental Users in Collaborative Filtering Recommender Systems

Philipp Meister, Lukas Wegmeth, Tobias Vente and Joeran Beel

Robust Solutions for Ranking Variability in Recommender Systems

Bonifacio Marco Francomano and Fabrizio Silvestri

Greedy Ensemble Selection for Top-N Recommendations

Tobias Vente, Zainil Mehta, Lukas Wegmeth and Joeran Beel

Robust Training of Sequential Recommender Systems with Missing Input Data

Federico Siciliano, Shoval Lagziel, Iftah Gamzu and Gabriele Tolomei

Schedule

Start Time

End Time

Event

Presenter

14:15

Opening

14:15

14:30

The Role of Fake Users in Sequential Recommender Systems

Filippo Betello

14:30

14:45

Removing Bad Influence: Identifying and Pruning Detrimental Users in Collaborative Filtering Recommender Systems

Lukas Wegmeth

14:45

15:00

Greedy Ensemble Selection for Top-N Recommendations

Tobias Vente

15:00

15:15

Reducing Popularity Influence by Addressing Position Bias

Andrii Dzhoha

15:15

15:30

Addressing bias in Recommender Systems: A Case Study on Data Debiasing Techniques in Mobile Games

Maria Paskevich

15:30

15:45

Robust Training Objectives Improve Embedding-based Retrieval in Industrial Recommendation Systems

Matthew Kolodner

15:45

16:15

Coffee break

16:15

16:30

Robust Solutions for Ranking Variability in Recommender Systems

Bonifacio Marco Francomano

16:30

16:45

Robust Training of Sequential Recommender Systems with Missing Input Data

Gabriele Tolomei

16:45

17:45

Keynote: Recommender Systems Robustness: the Impact of Attacks on Accuracy and Beyond

Ludovico Boratto, Associate Professor of Computer Science at the University of Cagliari (Italy)

17:45

Closing remarks

There are no rows in this table

⁠

Keynote Speaker

⁠

⁠

Ludovico Boratto

Associate Professor of Computer Science

University of Cagliari (Italy)

Organizers

⁠

⁠

Valerio Guarrasi

Federico Siciliano

Fabrizio Silvestri

Want to print your doc?
This is not the way.

Try clicking the ⋯ next to your doc name or using a keyboard shortcut (

CtrlP

) instead.