The main objective is to develop methodologies based on Data Mining (DM), within the framework of Intelligent Data Analysis processes, to extract knowledge that allows us to create a set of rules that govern the cooperative operation of a set of meta-heuristics from a centralized approach for the resolution of problems. The following specific objectives are planned in order to achieve this overall objective: given that the set of rules that are obtained by controlling the operation of the hybrid cooperative system of metaheuristics is a set of fuzzy rules, and since it is also very common to find situations in which information and data to mine contain or are described in an imprecise way, the first specific objective is to advance research in the study and development of DM techniques in imperfect scenarios that quickly and efficiently obtain or generate this type of model. One of the current problems for this type of DM based planning is the needed time and effort to build the database for future mining and knowledge extraction processes. The second specific objective is directed at alleviating this problem, namely to apply a building and data extracting process based on active learning or any learning strategy that allows the model to be upgraded in real time. The third specific objective provides advances in these systems by seeking to apply protocols to the complete design process of hyper-heuristics in order to control the set of heuristic that make it up so that they are more efficient and can be used for more general purposes. Since the complete knowledge extraction process is applied to data with characterized by imperfections, we are faced with the problem of the scarce support that present day DSS tools offer in work related to this type of information. Hence we propose improvements in the stages of knowledge extraction processes under imperfect conditions that serve as a guide to design a DSS platform in this type of scenarios.