随着语义Web和知识图谱的发展，知识库技术受到学术界与业界的广泛关注。OWL是表达万维网上知识的一个标准语言，其逻辑基础是描述逻辑。OWL本体推理对于基于本体的知识服务，比如说基于本体的数据集成，具有重要的作用。但是，现有的OWL本体推理机在处理海量知识的推理时存在诸多缺陷，迫切需要新的并行推理理论以及实现高效并行OWL本体推理机。在大数据领域，图结构越来越成为大规模数据处理的首选结构。而且一些基于图的推理机显示出较好的推理效率。本项目研究基于图的并行OWL本体推理的理论和技术，处理海量知识的推理。项目首先研究OWL本体的图表示方法；其次，从计算复杂度角度研究OWL本体对于标准推理任务的可并行性理论。再次，研究基于图的并行本体推理算法和优化方法。最后，实现一个基于图的并行OWL本体推理平台。本项目的成果将应用于语义Web、自然语言理解和智能问答等领域。

With the development of the Semantic Web and Knowledge Graph, knowledge base technologies have attracted widely attention in academia and industry. OWL is a W3C recommended standard language for representing knowledge on the Web, with description logics as its logical underpinning. OWL ontology reasoning plays an important role in knowledge-based services, such as ontology-based data integration. However, current description logic reasoners have weakness in dealing with large scale knowledge. Thus, we need new theories and framework for parallel reasoning. In big data area, graph structure has become the primary structure for handling large-scale data and some graph-based reasoners have shown good performance. In this project, we will study theories and techniques for graph-based parallel OWL ontology reasoning, which can be used to deal with large scale knowledge bases represented by OWL. We will first study the graph representation methods for OWL ontologies. Then, we will do research on graph-based parallel reasoning.algorithm and optimization techniques. Based on the theoretical work and algorithms, we will implement a graph-based parallel OWL ontology reasoning platform. The results of this project will be applied to Semantic Web, natural language understanding and intelligent question answering.