在上下文本体模型中,根据现有上下文信息推导出新知识,但在推理过程中存在两个问题:(1)现有上下文中可能隐含多个有用信息,而现有方法在推理前并未对其针对这一点进行处理,上下文具有不完整性,推理出的知识可能不全面;(2)推理后有新知识出现,新知识与旧知识可能存在不协同等问题,使得本体可能没有较好的可扩展性.针对以上两个问题,借鉴粗糙FCA的粗糙处理方法,提出基于粗糙FCA上下文抽取方法以获得隐含上下文;再使用概念代数将得到的所有上下文深度形式化表示,并构建具有较好可扩展性的概念网.实验结果表明,在提出的方法基础上进行上下文推理的正确率高于直接使用原推理方法,而且在本体可扩展性方面有明显优势.%In the contextual ontology model,new knowledge is derived from existing contextual information,but there are two problems in reasoning:(1)The existing context may imply a lot of useful information,but the existing method does not deal with this point before the reasoning, the context is incomplete and the reasoned knowledge may not be comprehensive.(2)After the reasoning, some issues make the ontology may not have better scalability, such as Emergence of new knowledge;new knowledge and old knowledge are not coordinated and so on.Aiming at the above two problems,this paper draw on the rough processing method of rough FCA, and proposed to extract the implicit context based on coarse FCA context extraction method.Then the concept algebra was used to formally represent the depth of all the obtained contexts and built a concept network with better scalability.Experimental results showed that the accuracy of context inference based on the proposed method was higher than that of direct inference,and it had obvious advantages in ontology scalability.
展开▼
