An artificial intelligence-augmented multi-criteria decision framework for identifying critical determinants of anti-immigrant attitudes in societal systems

Olgun E., Eti S., Çetinkaya H., Yüksel S., Dinçer H.

Social Sciences and Humanities Open, cilt.1, sa.1, ss.1-25, 2026 (Scopus)

Özet

Anti-immigrant attitudes define prejudiced, exclusionary, and discriminatory approaches toward immigrants, and these attitudes have multilayered economic, psychological, and social consequences that threaten social cohesion in societies. The fundamental problem of this study is that, despite the multidimensional nature of the factors fueling anti-immigrant sentiment, there is no clear consensus in the literature on which of these factors are most decisive. This deficiency leads to the ineffectiveness or incompleteness of intervention strategies developed to address anti-immigrant sentiment. Therefore, it is vital to analyze the causes of anti-immigrant attitudes with a more systematic and prioritized approach. In this context, the aim of this study is to identify the most critical factors influencing anti-immigrant attitudes and to develop an innovative fuzzy multi-criteria decision-making model for this purpose. The scarcity of systematic studies in the literature that rank the relative importance of these factors and the fact that existing studies rely on homogeneous expert opinions form the basis of this research. This study analyzes 11 criteria categorized under economic, cultural, psychological, and social headings. Expert opinions are evaluated using a machine learning-based expert weighting technique, considering demographic characteristics (level of education, experience, etc.). Fractal fuzzy sets are used to model decision-makers' uncertainties more flexibly and realistically, and criterion weights are calculated using the CIMAS (Criteria Impact–Loss Assessment Strategy) method. This unique model contributes to the literature in three ways: (1) Fractal fuzzy sets provide greater representativeness and modeling flexibility; (2) Machine learning-based expert weighting reduces subjectivity and increases the realism of the decision-making process; and (3) The CIMAS method provides a multidimensional and reliable assessment by considering the interactions and losses among the criteria. According to the findings, the populist rhetoric of political actors stands out as the most important factor influencing anti-immigrant attitudes. Furthermore, media representations of immigrants and economic crises also play a decisive role.

Anti-immigrant attitudes define prejudiced, exclusionary, and discriminatory approaches toward immigrants, and these attitudes have multilayered economic, psychological, and social consequences that threaten social cohesion in societies. The fundamental problem of this study is that, despite the multidimensional nature of the factors fueling anti-immigrant sentiment, there is no clear consensus in the literature on which of these factors are most decisive. This deficiency leads to the ineffectiveness or incompleteness of intervention strategies developed to address anti-immigrant sentiment. Therefore, it is vital to analyze the causes of anti-immigrant attitudes with a more systematic and prioritized approach. In this context, the aim of this study is to identify the most critical factors influencing anti-immigrant attitudes and to develop an innovative fuzzy multi-criteria decision-making model for this purpose. The scarcity of systematic studies in the literature that rank the relative importance of these factors and the fact that existing studies rely on homogeneous expert opinions form the basis of this research. This study analyzes 11 criteria categorized under economic, cultural, psychological, and social headings. Expert opinions are evaluated using a machine learning-based expert weighting technique, considering demographic characteristics (level of education, experience, etc.). Fractal fuzzy sets are used to model decision-makers' uncertainties more flexibly and realistically, and criterion weights are calculated using the CIMAS (Criteria Impact–Loss Assessment Strategy) method. This unique model contributes to the literature in three ways: (1) Fractal fuzzy sets provide greater representativeness and modeling flexibility; (2) Machine learning-based expert weighting reduces subjectivity and increases the realism of the decision-making process; and (3) The CIMAS method provides a multidimensional and reliable assessment by considering the interactions and losses among the criteria. According to the findings, the populist rhetoric of political actors stands out as the most important factor influencing anti-immigrant attitudes. Furthermore, media representations of immigrants and economic crises also play a decisive role.