基于数据挖掘的本体构建与重构技术研究

本体理论在知识工程领域得到广泛关注和普遍认可,构建完备且准确的领域本体已经越来越重要,同时,企业知识资源的更新与集成要求本体的不断进化与融合；针对目前本体构建与重构过程中数据处理效率低的问题,运用支持向量机分类及K-均值聚类的方法对本体构建数据进行处理,从文本数据中抽取关注的特定的信息,运用基于二叉树的多分类支持向量机以及支持向量机与K-均值融合的多样本聚类,总结基于分类与聚类的本体构建过程,并以离散型和连续型两种数据样本验证了方法的可行性；最后,在上述框架与理论研究的基础上,设计并开发了面向知识管理的本体工具平台,简单介绍系统的模块功能；实验结果表明,基于数据挖掘的本体构建与重构技术具有良好的应用效果。

Researches on Ontology Construction and Reconstruction Based on Data Mining

At present, ontology theory has attracted wide attention in the field of knowledge engineering. The construction of prefect and accurate domain ontology is getting more and more important, and at the same time, the update and integration of enterprise knowledge resource requires incessant evolution and merging of ontology. Aiming at the situation that process efficiencies and ontology integration is too slow, we use support vector machine classification and K- means clustering method to construct data processing. The thesis obtained specific information from the text data, and presented multiple-classification SVM and K-means clustering. Then, classification and clustering process was concluded for ontology construction and reconstruction, taking both discrete and continuous data sample as testing cases. The experimental results show that the proposed based on the ontology construction and reconstruction of data mining technology has good application effect.