Technology Category
Data Mining & Knowledge Management: Recommender SystemsMachine Learning: Learning Preferences or RankingsHumans and AI: Learning Human Values and Preferences
Which books do I like?
🤖 AI Summary
Readers often struggle to articulate their literary preferences, and niche reading interests are frequently overlooked by mainstream recommender systems. Method: This paper proposes ISAAC—a closed-loop human-AI co-creation framework integrating introspective support, AI-powered annotation, and curatorial guidance. Grounded in user book reviews, ISAAC employs LLM-driven fine-grained semantic book annotation, feedback-guided preference pattern identification, preference-enhanced collaborative filtering, and explainable recommendation algorithms to enable personalized, customizable, and traceable literary preference modeling. Contribution/Results: Experiments demonstrate that ISAAC significantly improves recommendation accuracy and deepens users’ self-reflection, effectively uncovering individualized reading structures. Its novel dual-track mechanism—automated annotation coupled with subjective introspection—overcomes longstanding limitations in explainability and niche coverage inherent in conventional recommenders. Moreover, ISAAC surfaces critical boundary issues, including data bias and narrative misinterpretation, thereby advancing both practical recommendation quality and theoretical understanding of literary preference formation.
Technology Category
Application Category
📝 Abstract
Finding enjoyable fiction books can be challenging, partly because stories are multi-faceted and one's own literary taste might be difficult to ascertain. Here, we introduce the ISAAC method (Introspection-Support, AI-Annotation, and Curation), a pipeline which supports fiction readers in gaining awareness of their literary preferences and finding enjoyable books. ISAAC consists of four steps: a user supplies book ratings, an AI agent researches and annotates the provided books, patterns in book enjoyment are reviewed by the user, and the AI agent recommends new books. In this proof-of-concept self-study, the authors test whether ISAAC can highlight idiosyncratic patterns in their book enjoyment, spark a deeper reflection about their literary tastes, and make accurate, personalized recommendations of enjoyable books and underexplored literary niches. Results highlight substantial advantages of ISAAC over existing methods such as an integration of automation and intuition, accurate and customizable annotations, and explainable book recommendations. Observed disadvantages are that ISAAC's outputs can elicit false self-narratives (if statistical patterns are taken at face value), that books cannot be annotated if their online documentation is lacking, and that people who are new to reading have to rely on assumed book ratings or movie ratings to power the ISAAC pipeline. We discuss additional opportunities of ISAAC-style book annotations for the study of literary trends, and the scientific classification of books and readers.
Problem
Research questions and friction points this paper is trying to address.
Identifying personal literary preferences and enjoyable books.
Developing a method combining AI and user introspection for book recommendations.
Addressing challenges in book annotation and recommendation accuracy.
Innovation
Methods, ideas, or system contributions that make the work stand out.
AI-annotated book recommendations
User-driven literary preference analysis
Explainable and customizable book suggestions
🔎 Similar Papers
2024-05-09arXiv.orgCitations: 4
