GENERATION OF PERSONALIZED READING RECOMMENDATIONS WITH ARTIFICIAL INTELLIGENCE INTEGRATION: DEVELOPMENT OF THE BOOKSUGGEST AI APPLICATION

Abstract

Recommendation systems have become central elements in contemporary digital platforms, thereby assisting users in making decisions related to content consumption. However, reading-based recommendations still have strong characteristics such as dependence on collective evaluations, popularity algorithms, or superficial metadata. This article presents a new possibility: BookSuggest AI, a recommendation system that integrates personal reading history, recorded by the user in Google Sheets spreadsheets, with Generative Artificial Intelligence models. Based on classic authors of recommendation systems, such as Adomavicius and Tuzhilin (2005), Goldberg et al. (1992), and Resnick and Varian (1997), the work explores how personal data can be transformed into relevant recommendations using modern AI techniques. The study details the technological architecture, the authentication process via Google OAuth, the data extraction and processing pipeline, as well as the use of generative models to build recommendations. The results demonstrate that BookSuggest AI is capable of generating personalized suggestions that are justified and consistent with user preferences. In addition, a critical analysis of the system and a discussion of its potential, limitations, and contributions are included. The article complies with ABNT standards and the editorial standards of Revista Aracê.