Entrepreneurial competencies and the conceptual dimension of intentions: Insights from a hybrid machine learning approach

PURPOSE: This study examines the relationship between entrepreneurial competencies and intention-related characteristics among university students, addressing the need for a deeper understanding of how psychological traits shape entrepreneurial tendencies in young populations. Focusing on students at the University of Information Technology and Management (UITM) in Poland, the research explores how adaptability, problem-solving, and cognitive flexibility contribute to the conceptual configuration of entrepreneurial intention. METHODOLOGY: A hybrid methodological approach was adopted, combining bibliometric analysis using SciMAT with data-mining techniques. A survey of 1,520 students provided the empirical basis for the analysis. Rough set theory was used to address incomplete data, while the C5.0 decision-tree classifier and feature interdependency analysis were applied to identify informative item-level patterns in the dataset. Demographic variables were incorporated to examine group-differentiating structural patterns across student groups. The study draws on TPB and self-efficacy as theoretical lenses for interpreting the conceptual patterns identified, without modelling TPB constructs as predictive variables. FINDINGS: The analysis highlights that adaptability (A12), problem-solving ability (A19), goal orientation (A18), and cognitive flexibility (A13) recur in the most informative branches of the decision-tree structures, indicating their central role within the broader configuration of entrepreneurial competencies. The results reflect how these attributes cluster within intention-related patterns rather than forming predictive relationships. The study also reveals distinct competency profiles across gender, age, nationality, and field of study, underscoring the heterogeneous nature of entrepreneurial characteristics in the student population. IMPLICATIONS: The findings contribute to the entrepreneurship literature by demonstrating conceptual coherence between intention-related attributes observed at the item level and dominant thematic patterns identified in recent research. Rather than testing or extending formal intention theories, the study offers an interpretative perspective on how adaptability, resilience, and problem-solving attributes cluster within student populations. From a practical standpoint, the results highlight the importance of entrepreneurship education initiatives that foster adaptive learning, coping with difficulty, and problem-solving skills. The use of data-driven decision-support tools may further assist educators in designing personalized learning environments that respond to heterogeneous student profiles. ORIGINALITY/VALUE: This study offers a novel contribution by integrating bibliometric validation with machine-learning-based pattern discovery. By mapping the conceptual landscape of intention-related attributes rather than predicting entrepreneurial intention, it provides a distinctive analytical perspective and actionable insights for educators and policymakers seeking to cultivate entrepreneurial competencies among university students.