基于本体视图特征项抽取方法研究*

为提供比单纯词汇信息更高效的概念特征信息和深层语义信息，并满足面向同一文本的多检索需求，在半自动化智能检索框架中引入本体视图，提出一种基于本体视图的特征项抽取方法。此方法首先针对文本特征建立本体视图；然后结合文本信息进行特征项抽取和类型映射，得到特征项集；最后基于特征项集完成检索过程。检索结果显示，基于本体视图特征项抽取方法能改善检索系统的性能，提高检索的准确率和效率。

Study on feature item extraction method based on ontology view

This paper proposed a feature item extraction method based on ontology view by the introduction of ontology view in semi-automatic intelligent information retrieval framework. In order to provide concept feature and implied semantic which were more efficient than vocabulary information, and satisfied the demand for multi-queries faced the same Chinese text. Firstly, this method generated ontology view for text characteristic, and then obtained the set of feature item by extracted and mapped feature items with Chinese text information. At last, finished the process of multi-queries based on the set of feature item. The results show that the feature item extraction method based on ontology view can improve the performance of retrieval systems and improve the precision and efficiency of intelligent retrieval.