基于启发式规则的自动化本体扩充

自动化地获取网络资源中的领域本体可以缩短本体的构建周期,但自动化的本体扩充还是本体工程中的一个挑战,其难点主要在于如何抽取术语并在新术语和已有本体之间建立映射关系。为此,提出了一个基于启发式规则的本体自动化扩充方法。该方法从网络资源中抽取自然语言文本,结合自然语言处理技术进行文本预处理,采用优先匹配对象属性的方式挖掘领域知识术语,然后通过启发式规则匹配术语的方式进行本体扩充,最后进行一致性检测。采用上述方法实现了一个基于Web的本体扩充工具。以城市景观信息核心本体作为研究案例进行了实验,结果显示本方法在扩充实例时具有较高的查准率和查全率,表明其具有有效性和可行性。

Automatic Ontology Population Based on Heuristic Rules

The cycle of building ontology can be shortened by means of automatically extracting domain ontology in Internet resources,but automatic ontology population is still a challenge in ontology engineering.There are two difficulties in this area,which are how to extract terms and how to construct the mapping relationship between the new terms and the existed ontology.Therefore,this paper proposed a method for automatic ontology population based on the proposed heuristic rules.This method extracts natural language texts from the Internet,combines traditional natural language processing methods for text preprocessing,discovers domain terms by preferentially matching object properties,enriches the ontology by matching these terms using heuristic rules,and finally checks the consistency of the enriched ontology.On the base of the proposed method,this paper also implemented a Web-based tool for ontology population.Using an urban landscape information core ontology as a case study,the experimental results show that the method for enriching ontology individuals has a high precision and recall.The results also prove that the proposed method is effective and feasible.