The motivation behind this paper is to propose a new approach to reformulate the objective function based on the decomposition algorithm in an interactive model for finding the optimal solution to Multi-Level Large Scale Quadratic Programming Problem (MLLSQPP), in which the objective function can be factorized into two single linear objective functions and has block angular structure of the constraints. Also, a concepts of satisfactoriness as the upper level decision-makers' preferences are advanced until the preferred solution is obtained. The bi-section method is suggested as an essential solution method to assess these upper level decision-makers' preferences precisely. The adequacy of this proposed approach is illustrated through comparing with the Taylor's polynomial series approach to prove the applicability of this proposed approach for solving this particular type of (MLLSQPP). Finally, in order to demonstrate the proficiency of the proposed model, an illustrative example is provided.