Numerous studies indicate that current recommender systems primarily focus on customer satisfaction, dissatisfaction, and personalized preferences when making product recommendations. However, these systems often neglect the anxiety customers may feel when choosing between similar products. This unease can result in poor decision-making and suboptimal choices. The ideal scenario for customers is to select a product without experiencing anxiety. Our study addresses this gap by incorporating "tranquillity" (or anxiety) as a behavioral factor in the recommendation process. Failing to consider these intuitive customer judgments can lead to the selection of inappropriate products. We propose a unified personalized recommendation approach using interval- valued intuitionistic fuzzy sets, which accounts for uncertain, conflicting criteria and customer behavior. This methodology identifies the best alternative by considering the customer's flexible preferences through an averaging operator. We compare the effectiveness of our approach with existing studies and demonstrate its applicability using a car purchase example in e-commerce