Reimagining Social Robots as Recommender Systems: Foundations, Framework, and Applications

TL;DR

This work addresses the limited ability of current social robots to personalized interactions across long-term, short-term, and fine-grained user preferences. It proposes a principled integration of Recommender Systems (RS) into the social-robot pipeline, implemented as modular components for user profiling, ranking, and responsible computing, and formalized through losses and scores such as $\mathcal{L}^\text{CF}$, $\mathcal{L}^\text{Seq}$, and $\hat{r}_{u,a}$. The authors detail a general framework with modules $\mathcal{P}$, $\mathcal{C}$, $\mathcal{A}$ and $\text{VLAM}$, and provide concrete techniques for long-term, short-term, and fine-grained preferences, plus a retrieval-and-ranking paradigm to efficiently select actions. They also address privacy, fairness, and transparency via federated learning, unlearning, IPS-based debiasing, and other responsible computing methods, and discuss actionable use cases and open challenges. The framework aims to accelerate collaboration between RS and HRI communities by offering plug-and-play RS components that enhance personalization while preserving safety and ethical standards in real-world social-robot applications.

Abstract

Personalization in social robots refers to the ability of the robot to meet the needs and/or preferences of an individual user. Existing approaches typically rely on large language models (LLMs) to generate context-aware responses based on user metadata and historical interactions or on adaptive methods such as reinforcement learning (RL) to learn from users' immediate reactions in real time. However, these approaches fall short of comprehensively capturing user preferences-including long-term, short-term, and fine-grained aspects-, and of using them to rank and select actions, proactively personalize interactions, and ensure ethically responsible adaptations. To address the limitations, we propose drawing on recommender systems (RSs), which specialize in modeling user preferences and providing personalized recommendations. To ensure the integration of RS techniques is well-grounded and seamless throughout the social robot pipeline, we (i) align the paradigms underlying social robots and RSs, (ii) identify key techniques that can enhance personalization in social robots, and (iii) design them as modular, plug-and-play components. This work not only establishes a framework for integrating RS techniques into social robots but also opens a pathway for deep collaboration between the RS and HRI communities, accelerating innovation in both fields.

Figures (2)

• Figure 1: The typical HRI flow with plug-and-play RS modules. The RS modules enhance robot cognition through comprehensive user profile modeling, optimal action ranking, and responsible computing, mapping perceptual signals into personalized response choices.
• Figure 2: RS-augmented HRI decision loop in action selection and response generation.