Abstract: RDF is a Web data model proposed by W3C, RDF stream is a data model that complies with the Linked Open Data and extends RDF, and it is a kind of dynamic RDF data set with infinite growth over time. In RDF query processing, the RDF pattern is the pattern of static scheduling execution plan, which cannot meet the characteristics of dynamic RDF stream in real time. Therefore, an ant colony optimization approach for reordering RDF stream pattern in RDF stream query processing is proposed. The main research contents include：A RDF stream cost evaluation model is proposed Based on the AND-OR directed graph; Max-Min ant colony optimization algorithm is improved, and uses improved algorithms for reordering RDF stream pattern. It is observed that the proposed approach can reorder RDF stream execution plan efficiently and dynamically.

Key words: RDF stream pattern, Max-Min ant colony optimization algorithm, Resource Description Framework（RDF）, AND-OR directed graph

摘要： RDF是由W3C提出的一种Web数据模型，RDF流是一种遵从链接开放数据（Linked Open Data）、扩展RDF的数据模型，是一种随时间延续无限增长的动态RDF数据集合。在RDF查询处理中，RDF模式是静态编排执行策略的模式，无法满足动态RDF流实时变化的特性。因此，提出一种在RDF流查询处理中重排序RDF流模式的蚁群优化方法。主要研究内容包括：提出了基于AND-OR有向图RDF流代价评估模型；改进了Max-Min蚁群优化算法，并使用改进算法重排序RDF流模式。实验表明，该方法能够高效地动态编排RDF流执行策略。

关键词: RDF流模式, Max-Min蚁群优化算法, 资源描述框架（RDF）, AND-OR有向图