Recommendation systems have applications in many fields of e-commerce and social networking platforms, helping to bring products closer to potential customers and enabling manufacturers to introduce more products to consumers. An important aspect is the group recommendations. Unlike traditional systems that focus on individuals, group recommendation aims to serve multiple users simultaneously. If the weights of individuals are not provided in advance, they are often assigned randomly or based on heuristics, which is often difficult to prove as meaningful and correct. In this study, we propose a novel and effective approach for multidimensional vector ranking, which arranges weights for individuals in a group based on their product evaluation history. The proposed method is implemented, compared with traditional techniques, and evaluated for statistical significance.