Overview

Recommender systems are a popular kind of information access systems that provides personalized recommendations to users based on their preferences (i.e., current and past), behaviors, and feedback. These systems are widely used in e-commerce, social networking, and content sharing platforms, where the volume of available data and options can be overwhelming for users. Recommender systems help users discover new items, products, or content that match their interests and needs, and enhance their overall experience by saving time, effort, and money. Among the different aspects involved in the relationship between humans and recommender systems, User-centered design plays a core role. This approach involves users in many aspects of the design process to stress trust, transparency, and efficacy as key factors for the successful adoption and acceptance of recommender systems. In order to obtain these results, the integration of recommender systems with interfaces designed from users’ perspectives is crucial. Interfaces that are user-friendly, intuitive, and visually appealing can facilitate the interaction between users and recommender systems, and increase the perceived usefulness and credibility of the recommendations. Moreover, interfaces that allow users to provide feedback, adjust their preferences, and control the level of personalization can increase their engagement and satisfaction with the system. One of the key challenges in designing interfaces for recommender systems is to balance the level of personalization with the diversity and serendipity of the recommendations. While users may expect the system to recommend items that match their exact preferences and past behaviors, too much personalization can lead to a narrow and repetitive experience that limits their exploration and discovery of new options. Therefore, interfaces that provide a variety of options, alternative recommendations, and serendipitous discoveries can create a more engaging and rewarding experience for users.

In recent years, the focus has shifted from pure AI-based recommender systems to human-centered AI, where the emphasis is on involving human users in the design, evaluation, and integration of these systems. Human-AI collaboration and human-centered AI are emerging research areas that focus on designing systems that work in tandem with human users rather than replacing them. There is a growing interest in designing systems that are inclusive and respectful of diverse user needs and preferences, including cultural, social, and individual factors. This approach can lead to more inclusive recommendations that better reflect the diversity of users. Human-AI collaboration and human-centered AI within recommender systems are focused on designing systems that work with and for humans improving user satisfaction and trust. Recent research on human-AI collaboration involves several critical areas of investigation, such as Human-in-the-loop, Symbiotic AI, Explainable AI, User-centered design, and Intelligent Interfaces. Overall, this area of research is aimed at developing systems that can work effectively with human users, considering their preferences, cognitive abilities, and ethical values. They should be transparent, interpretable, adaptable, and respectful of the user’s autonomy and privacy. The ultimate goal is to develop recommender systems that can support the user’s decision-making process, enhance their well-being, and promote social good.

The IntRS workshop series focuses on user-centric perspective on recommender systems research. The IntRS workshop brings together an interdisciplinary community of researchers and practitioners who share research on new recommender systems (informed by psychology), including new design technologies and evaluation methodologies, and aim to identify critical challenges and emerging topics in the field. Indeed, the workshop focuses particularly on the impact of interfaces on decision support and overall satisfaction, and it is also connected to the topics of Human-Centered AI, Explainability of decision-making models, User-adaptive XAI systems, which are becoming more and more popular in the last years especially in domains where recommended options might have ethical and legal impacts on users. The integration of XAI with recommender systems is crucial for enhancing their transparency, interpretability, and accountability. XAI can help users understand why a particular recommendation is made, what data and algorithms are used, and what factors influence the outcome. This can increase the user’s trust and confidence in the system, and improve their satisfaction and engagement with the recommendations. The explanations should be presented in a way that is understandable, concise, and relevant to the user’s context and goals. This requires collaboration between XAI researchers, designers, and end-users to ensure that the explanations meet the user’s expectations and needs. An interesting research direction that has recently received renewed interest is to investigate how users interact with recommenders based upon their cognitive model of the system. Previous work, investigated the impact of users’ mental models of recommender systems on their interactions with them and drew a theory to understand the key determinants motivating users to such user behavior. We believe that the paradigm that describes the relationship between humans and recommender systems is changing and evolving from a “human-centered” design approach toward a symbiotic vision. From this point of view, the mutual exchange of knowledge between human and system will lead us towards “symbiotic recommender systems”, in which both parties learn by observing each other. We hope IntRS will be the forum where fresh ideas on this topic will be discussed.