【摘要】 结合关系模型、产生式规则和自主计算三者的优点,提出了一个任务驱动的自主推理的动态领域知识应用模型。在该模型中,将领域知识按需要进行聚类,根据任务进行规则映射和推理模块中不同处理器的选择,运用证据组合方法合并各处理阶段的不确定度,采用基于自适应控制策略的自主计算技术进行知识推理。推理结果被应用到不确定模式集成模型的其他模块中,辅助解决模式集成中存在的不确定性问题。经实例验证该模型是可行的。更多还原

【Abstract】 The dynamic Domain Knowledge Application Mode(lDKAM) based on tasks is proposed.Automatic reasoning can be finished in the model with three advantages of the relational model,production rules,and autonomic computing.Domain knowledge is clustered according to needs.The mapping operation of rule and the choice operation of different processor are executed according to tasks.Evidence combination method is applied to combine uncertain degree of each processing stages.Autonomic computing technology based on self adaptive control strategies is adopted during knowledge reasoning process.Rea-soning results are applied in other models of Uncertain Schema Integration Mode(lUSIM).DKAM can assist to resolve existing uncertain problems in schema integration.The example shows that DKAM is feasible.更多还原