Search for: All records

Generative AI, particularly Large Language Models (LLMs), has revolutionized human-computer interaction by enabling the generation of nuanced, human-like text. This presents new opportunities, especially in enhancing explainability for AI systems like recommender systems, a crucial factor for fostering user trust and engagement. LLM-powered AI-Chatbots can be leveraged to provide personalized explanations for recommendations. Although users often find these chatbot explanations helpful, they may not fully comprehend the content. Our research focuses on assessing how well users comprehend these explanations and identifying gaps in understanding. We also explore the key behavioral differences between users who effectively understand AI-generated explanations and those who do not. We designed a three-phase user study with 17 participants to explore these dynamics. The findings indicate that the clarity and usefulness of the explanations are contingent on the user asking relevant follow-up questions and having a motivation to learn. Comprehension also varies significantly based on users’ educational backgrounds. 

This paper addresses a gap in the AI governance literature in understanding collaboration between national governments and tribal nations in governing AI systems for emergency management. This conceptual work develops and presents a governance design framework for accountable AI systems to fill the knowledge gap by drawing from the fields of public administration, information systems, indigenous studies, and emergency management. This framework situates the governance framework in a cross-sovereignty historical, legal, and policy contexts. It captures the multi-level features and embeddedness of governance structures, including the levels of collaborative governance structure, software system governance rules, and technical software system design. The focal governance dynamics involve the collaborative process in the bi-directional relationship between governance rules and technical design for accountability and the feedback loop. The framework highlights the importance of multi-level and process considerations in designing accountable AI systems. Productive future research avenues include empirical investigation and resulting refinement of the framework and analytical rigor employing institutional grammar. 

Tribal governments bear an uneven burden in the face of escalating disaster risks, climate change, and environmental degradation, primarily due to their deeply entrenched ties to the environment and its resources. Regrettably, accessing vital information and evidence to secure adequate funding or support poses difficulties for enrolled tribal members and their lands. In response to these challenges, this paper collaborates with tribal nations to co-design intelligent disaster management systems using AI chatbots and drone technologies. The primary objective is to explore how tribal governments perceive and experience these emerging technologies in the realm of disaster reporting practices. This paper presents participatory design studies. First, we interviewed seasoned first-line emergency managers and hosted an in-person design workshop to introduce theEmergency Reporterchatbot. Second, we organized a follow-up design workshop on tribal land to deliberate the utilization of drones within their community. Through qualitative analysis, we unveiled key themes surrounding integrating these emergency technologies within the jurisdiction of tribal governments. The findings disclosed substantial backing from tribal governments and their tribal members for the proposed technologies. Moreover, we delved into the potential of chatbots and drones to empower tribal governments in disaster management, safeguard their sovereignty, and facilitate collaboration with other agencies. 

This paper introduces an innovative approach to recommender systems through the development of an explainable architecture that leverages large language models (LLMs) and prompt engineering to provide natural language explanations. Traditional recommender systems often fall short in offering personalized, transparent explanations, particularly for users with varying levels of digital literacy. Focusing on the Advisor Recommender System, our proposed system integrates the conversational capabilities of modern AI to deliver clear, context-aware explanations for its recommendations. This research addresses key questions regarding the incorporation of LLMs into social recommender systems, the impact of natural language explanations on user perception, and the specific informational needs users prioritize in such interactions. A pilot study with 11 participants reveals insights into the system’s usability and the effectiveness of explanation clarity. Our study contributes to the broader human-AI interaction literature by outlining a novel system architecture, identifying user interaction patterns, and suggesting directions for future enhancements to improve decision-making processes in AI-driven recommendations.