基于改进NSGA-Ⅱ算法的科技服务链推荐

Science and Technology Service Chain Recommendation Based on Improved NSGA-II Algorithm

摘要:

在科技服务机构和科技服务大量增长以及用户需求日趋复杂化和链式化的背景下，向用户推荐满足其个性化要求的科技服务链成为亟需解决的问题。本文首先构建考虑服务特有属性的科技服务推荐模型，在NSGA-II算法中引入去重操作以消除由于候选服务数量限制产生的重复服务链，提高算法的多样性和收敛性，并使用该算法对模型进行求解，得到一组Pareto最优服务链集。然后通过科技服务链综合评估函数对服务链进行排序，将最优服务链推荐给用户。实验结果表明改进的NSGA-II算法求解出的满足服务需求方要求的解的数量和准确性均优于NSGA-II算法。

关键词:科技服务;服务链推荐; NSGA-II ;服务质量

Abstract:

Under the background of the massive growth of science and technology service institutions and science and technology service as well as the needs of users become increasing complicated and more like a chain, it has become an urgent problem to recommend the science and technology service chain to users to meet their personalized requirements. In this paper, a technology service recommendation model considering the unique attributes of science and technology services is first constructed, then a deduplication operation is introduced in NSGA-II algorithm to eliminate the repeated service chain caused by the limit of the number of candidate services and improve the diversity and convergence of the algorithm.And this algorithm is used to solve the model and a set of Pareto service chains is obtained.Finally, the service chain is sorted by the comprehensive evaluation function of science and technology service chain, and the optimal service chain is recommended to users. The experimental results show that the number and accuracy of the solutions that meet the requirements of service demands obtained by the improved NSGA-II algorithm are superior to that of the NSGA-II algorithm.

Key words:science and technology service; service chain recommendation; NSGA-II;quality of service