基于非结构化数据的本体学习研究

语义Web的创建需要一套共同的标准概念体系，即本体（Ontology）。而现在本体的构造手段仍然是以手工构造为主，效率和准确率都非常低，很容易导致知识获取的瓶颈。近年来，自动创建领域本体可以克服手工方法的不足，成为当前的研究热点之一；本体学习是自动或半自动构建本体的一系列方法和技术。提出了一种利用知网，基于非结构化数据的特定领域概念及其之间关系的提取算法，从军事领域选取4个种子概念：舰、导弹、机和炮，并通过实验测试了该算法。

Research of ontology learning based on non-structured data

Establishment of semantic Web requires a set of common standard concept system,namely ontology.Manual ontology construction has low efficiency and accuracy now which can easily result in a knowledge acquisition bottleneck.Overcoming the deficiency of the manual methods,the semi-automatic or automatic methods of building ontology are becoming the hotspot of current research recently,namely,ontology learning.Ontology learning is the set of methods and techniques used for building ontology in an automatic or semi-automatic fashion using several sources.A method is presented for discovering domain-specific concepts and relationships based on non-structured data through using HowNet in this paper.The method is tested on four seed concepts selected from the martial domain: warship,missile,plane,and gun.