论文部分内容阅读
有一种服务冲突来源于服务质量属性间的彼此矛盾,即某个质量属性的改善同时可能会导致另一属性的性能下降,从而成为了提高服务质量的障碍。本文借鉴TRIZ冲突矩阵的思路构建了这类服务冲突矩阵,并利用BP神经网络,把服务冲突矩阵变换为一个以服务参数为输入层,冲突解决原理为输出层,并包含一个隐含层的三层BP神经网络冲突消解模型。并以服务领域已解决的冲突问题为样本,对该模型进行了训练和检验,运行结果验证了该网络的可行性。
One conflict of service originates from conflicting quality of service attributes, that is, an improvement in one quality attribute may simultaneously lead to a decrease in the performance of another attribute, thereby becoming an obstacle to improving service quality. This dissertation constructs such service conflict matrix by referring to the TRIZ conflict matrix, and uses BP neural network to transform the service conflict matrix into an input layer with service parameters as input layer, the conflict resolution principle as output layer and an implicit layer of three Conflict Resolution Model Based on BP Neural Network. Taking the conflict problem solved in the service area as a sample, the model was trained and tested, and the operation result verified the feasibility of the network.