Modelo de sistema inteligente autónomo de compra en un contexto de economía social

This study designs, implements, and evaluates a generic shopping context model to recognize the effects of emotions and feelings on purchasing decisions. A decision-making model is developed that incorporates a sentimental factor related to altruism and is integrated into a rational behavioral framework. The system is evaluated through agent-based simulations to analyze its behavior and contribute evidence on the affective influence in artificial economic environments. This research lies at the intersection of economics and computational psychology. The results are analyzed through three contrasting scenarios. In the first case, composed exclusively of selfish agents, the system rapidly degrades: stock is depleted, consumption satisfaction and sentimental satisfaction drop sharply, and significant inequality arises. This configuration demonstrates that, without cooperative behavior, system sustainability is compromised from the early stages. In the second case, the inclusion of conflicted altruistic agents temporarily improves the system’s sustainability. Stock is preserved for more iterations, overall satisfaction remains high for longer, and agents’ sentimental responses adjust to the context. However, under severe scarcity, selfish behavior ultimately dominates, diluting the effects of altruism. This scenario shows greater resilience than the first, though it does not prevent the system’s programmed collapse. In the third case, generative altruistic agents enable more efficient stock accumulation in abundant contexts, extending system stability. Nevertheless, under critical conditions, these agents show lower adaptability compared to conflicted altruists. While achieving better initial performance than the first case, they fail to maintain long-term equity and satisfaction as observed in the second. This suggests that generative altruism is more effective in favorable environments but loses effectiveness when resources become scarce.