A classification model for patenting motivation: evidence from Brazilian public universities

Purpose This study aims to propose and validate a classification model to identify the motivational profiles of academic inventors. The research addresses the unique context of Brazilian public universities within an emerging economy, where these institutions are the primary drivers of national patenting activity. Design/methodology/approach Adopting a mixed-methods approach, this study combines a systematic literature review to build a novel classification framework with quantitative analysis of data from 450 serial inventors from Brazil’s 12 most prolific patenting universities. Data was collected via web scraping of the Lattes platform and a direct survey. A supervised machine learning algorithm (decision tree) was developed to automate profile classification and infer missing data for nonrespondents, ensuring methodological replicability. Findings The study yields two primary findings. Substantively, inventors are driven more by intrinsic challenges (“Puzzle”) than financial rewards (“Gold”), and a key paradox emerges: a strong entrepreneurial orientation coexists with scarce industry collaboration due to systemic barriers. Methodologically, the classification model reveals a significant self-selection bias, showing that survey respondents are far more intrinsically motivated than the inferred nonrespondent group. This quantifies the limits of relying solely on public data to predict complex human traits. Originality/value This paper’s value is a dual methodological contribution. It delivers a novel, automated framework for profiling innovator motivations and a transparent process for using machine learning to identify and correct for self-selection bias in social science surveys. By offering a replicable diagnostic tool while simultaneously demonstrating the boundaries of data-driven inference, the study presents a more rigorous approach to analyzing human capital in innovation. Its application provides critical, evidence-based insights for university managers and policymakers in Brazil and other emerging economies.