摘要
为提高本体在海量web中概念学习效果,提出基于仿射传播算法的本体学习框架,框架设计基于文本解析、相似度计算、概念抽取等过程获得概念列表.系统将抽取的语义和结构相似的术语作为数据点,通过算法迭代数据点之间的信息传递,直到生成高质量的样本集合,所有的样本集合组成学习的本体.实验证明,查准率和查全率都有较好的效果,方案具有可行性.
To improve the effect of ontology learning in a massive web,Ontology learning framework based on affinity propagation was proposed. Text analysis,similarity calculation,concept extraction were designed to get concept list. We present a approach for ontology learning,which takes as input semantic and structural similarity between pairs of extracted terms called data points. Messages are passed between data points until high quality set of exemplars emerges. All exemplars will be considered as domain concepts for learning domain ontologies. Results show that our approach achieves high precision and recall. So the program has good feasibility.
出处
《小型微型计算机系统》
CSCD
北大核心
2014年第7期1596-1598,共3页
Journal of Chinese Computer Systems
基金
国家自然科学基金项目(61005010)资助
安徽省高等学校省级优秀青年人才基金重点项目(2013SQRL074ZD)资助
安徽省自然科学基金面上项目(1408085MF135)资助
合肥学院科研发展重点基金项目(13KY02ZD)资助
关键词
本体学习
本体建设
领域概念
仿射传播
ontology learning
ontology construction
domain concepts
affinity propagation