Nowadays, the renaissance of data science and artificial intelligence (AI) has attracted huge attention from every corner of the world. Recommender Systems (RS), as one of the most popular and significant and practical applications of data science and AI, have been widely planted into our daily lives and has made a huge difference. In the current era of digital economics and information explosion, RSs can not only guide and change human’s behaviours, but also revolutionize human’s way of life.

With the advancement of data science and AI in recent decades, more and more powerful and accurate RSs have been developed to provide recommendation services in various areas including shopping, eating, travelling and entertainment. These RSs have achieved big success and benefit the society greatly. However, most of existing work on RS mainly focus on the improvement of recommendation accuracy, while ignoring other significant aspects, such as the trustworthiness (e.g., robustness, interpret-ability, transparency, security) of RS technique and models and the social impact (e.g., fairness among groups and individuals, influence on users’ behaviours) of recommendation results. These aspects are very important and cannot be overlooked since they determine whether the recommendation services are reliable, trustworthy and beneficial to both the end users and the society.

This workshop aims to engage with active researchers from recommendation communities and other communities including social science communities and deliver the state-of-theart research insights into the core challenges in providing responsible recommendation services. To be specific, we will focus on two main topics contributing to responsible RSs in this workshop:

(1) Developing reliable and trustworthy RS models and techniques to provide reliable (e.g., fair, interpretable, secure, responsible, etc.) recommendation results when facing a complex, uncertain and dynamic cyberspace which is full of threats from noise, attacks, and bias.
(2) Discovering the influence of recommender systems on human’s behaviours, recognition, etc. and ensuring the influence is positive and sustainable to the end users and the whole society. The workshop will provide an opportunity to promote the research on trustworthy RSs and the social impact of RSs, thus developing beneficial AI applications and better universalizing the advanced techniques to a wider range of the real life.

Topics of Interest
The workshop invites submissions on all topics of trustworthy RSs and/or social impact of recommendations, including but not limited to:

Fundamental or emerging data science or artificial intelligence theories, approaches and applications related to trustworthy recommendations
Social impact and influence of RSs on end users and the society
Mitigation of negative impact of RSs and enhancement of positive impact of RSs towards the sustainable development of community, society and culture
Recommendation with low-quality data, including highly sparse data, noisy or corrupted data, heavily  duplicated data, and biased data
Uncertainty modeling for recommendation where user interests frequently drift over time and/or results need to be presented in a highly dynamic environment
Robustness models for recommendations including attacks and counter approaches
Interpretable recommendation that provides persuasive explanations and/or generates faithful interpretations to the recommendation process
Fairness and debiasing, where a fair system is designed to balance its accuracy with potential biases and/or unfairness
Security and privacy-aware recommendations including federated recommendation, on-device training/inference, and privacy-protected ranking mechanisms
Human-in-the-loop computing for improving accuracy, explainability, or adaptivity
Surveys, evaluations, or benchmarking on state-of-the art research in the area of trustworthy recommendations
Novel and emerging applications of recommendation techniques, especially trustworthiness related approaches and solutions
Novel evaluation protocols, approaches and metrics for evaluating the trustworthiness of recommendations

Important dates
Workshop Paper Submission: May 26th, 2022
Workshop Paper Notification: June 20th, 2022
Workshop Date (tentative): August 14th or 18th, 2022

Enquiry
Please contact Dr. Shoujin Wang via shoujin.wang@mq.edu.au.