Study On Decision Making Methods For Reviewing And Selecting Scientific Fund Projects

In the knowledge economy era, the science and technology has been developing in the all-time speed. In such situation, every country makes effort on their technology polices in order to get advantages in such technology competition. The scientific fund is one of the best forms of supporting science and technology developments for governments. Therefore, many countries recognize the importance of implementing scientific fund, and pursue the improvement of scientific fund management level all the time. On the other hand, reviewing and selecting scientific fund projects rationally and fairly has been concerned by the scientific society. However, although many papers related with scientific fund can be found in the published literatures, there is still scarce of the systematic research. More over, some problems, to our knowledge, have not been stressed until now. Given such situation, this paper analyses the problem of reviewing and selecting scientific fund projects systemically, and then proposes several decision problems specifically, and finally proposes the corresponding decision making methods based on some decision theories like mathematical modeling methods and intelligent optimization algorithms. The main contributions of this paper are in the following:(1) Firstly, the methods for evaluating and selecting experts of scientific fund project are proposed. As for the expert evaluation, the evaluation criteria system is firstly constructed, and then the evaluation method is presented. As concerned the expert selection, two methods for selecting experts are proposed. In the first method, the expert evaluation value is stressed. Therefore, the expert with high evaluation value is more possible to be selected. In the second method, the experts'background is concerned. So, the diverse expert background is desired in this method. The proposed evaluation method is practicable and operable, and the proposed selection methods will remedy the gap in this respect of the current work.(2) Secondly, the method for grouping experts is proposed. The expert grouping problem, for the first time, is identified. After the analysis of the expert grouping problem, the expert ruler and corresponding measurement approach are discussed. Then, a mathematical model for expert grouping is formulated. Considering the characteristics of the model, a hybrid genetic grouping algorithm for expert grouping is designed and developed. Mass numerical experiments are conducted to verify the effectiveness of the proposed algorithm. Also, an numerical example is presented to illustrate the use of the proposed approach. The work of this paper on expert grouping problem, provide a new perspective for scientific fund management, which obviously will enrich the current related research.(3) Also, the method for proposal grouping is proposed. Firstly, the grouping rules are designed. Then, based on these rules, a multiobjective model for proposal grouping is built, in which the constraint is represented in a special way. To solve the model, a genetic algorithm is proposed. In summary, the proposed method for proposal grouping, will not only improve the effectiveness and efficiency of the proposal grouping, but also facilitate to the fair and objective project review.(4) Further, the methods for proposal assignment are proposed. Firstly, the research domain matching of proposal and experts are proposed and analyzed. Then, based on the work in above chapters, two proposal assignment methods are proposed. In the first one, the proposal is assigned to expert individually. In the second one, the proposal group is assigned to expert group. In each method, the corresponding assignment rules are set up, and the model with its solution method is described. Furthermore, the possible conflicts in assignment are analyzed, and several resolve rules are proposed. Finally, two assignment methods are compared to provide the application suggestions for the practice. In a word, the proposed assignment methods will further facilitate the effective and efficient proposal assignment.(5) Finally, a two-stage integration method for selecting projects is proposed. Firstly, four possible situations of the current review results are analyzed. Then, a framework of the two-stage integration method is built. After that, the two-stage integration decision making method for project selection is described in detail. The proposed method combines the evaluation information of both communication experts and session experts. Therefore, it will improve the current selection method.