Mental Health and Ambiguous Choice: A Big Data-Driven Decision Model with Interval Type-2 Fuzzy Sets
Hongyan LiAbstract
Objective
This study explores the intrinsic linkage between individual mental health states and uncertain decision-making behaviors. It constructs an integrated framework combining psychological well-being indicators and standardized preference models, aiming to clarify how mental fluctuations affect choices and improve the prediction accuracy of individual behavioral decision-making.
Subjects and Methods
A hybrid research design was implemented. First, public psychological indicators (e.g., collective anxiety, sentiment polarity) were extracted and quantified from extensive social media datasets using advanced Natural Language Processing (NLP) techniques. Second, a controlled experiment with 120 participants was conducted to elicit their risk preferences in ambiguous scenarios, which were mathematically modeled as IT2FPRs. A novel fusion model was then constructed to statistically couple the macro-level psychological data with the micro-level parameters of the individual fuzzy preference models.
Results
The analysis revealed a statistically significant correlation between specific public psychological patterns and the uncertainty footprints of individual IT2FPRs. For instance, periods of heightened collective anxiety were associated with wider uncertainty bounds in personal preference models, indicating increased decision hesitation. The proposed integrated framework demonstrated a 22% improvement in predicting real-world choice outcomes compared to models using only traditional preference or isolated sentiment analysis.
Conclusions
Dynamic mental health conditions and negative emotions like collective anxiety are key endogenous factors shaping risky and ambiguous decision preferences, where psychological fluctuations directly disrupt rational judgment. This big data-based model effectively reveals the coupling mechanism between mental health changes and uncertain decision-making and realizes quantitative identification of psychological pressure. It enhances behavioral prediction precision and provides technical support for public mental health monitoring, early warning and intervention, helping stabilize group mental states and optimize public mental health governance.
Acknowledgement
Anhui Provincial Department of Education Key Natural Science Project “Risk-Aware Interval Type-2 Fuzzy Preference Relations and Their Application in Health Management” (Project No: 2025AHGXZK31182).
Corresponding Author
Hongyan Li, School of Civil Engineering and Architecture, Anhui Communications Technical College, Hefei 230051, China.