摘要
描述了一个基于本体词汇的三维模型语义检索的方法。该方法首先对一个三维模型库的词汇进行语义上扩充,然后基于关键词进行检索,而不是做简单的文字匹配。其主要工作是利用WordNet英文本体和Jena推理引擎构建一个语义网络,然后在该语义网络上做推理查询,检索到相应的语义结果。实验结果表明,该方法能有效地将一个分类不够丰富的模型库的词汇做较大的扩充,使其覆盖的范围扩大,同时保证扩充的词汇和原有词汇在语义上的相关性,从而提高基于语义关键词检索三维模型的有效性。
This paper presented the construction of a 3D model retrieval system based on ontology vocabularies, which achieves 3D model retrieval using keywords based on a semantic vocabulary expansion on the model library, rather than simply literal comparison. A semantic network is firstly built by using WordNet （an English ontology） and Jena （an engine of reasoning）, by which the semantic retrieval results are then reasoned based on the constructed semantic net- work. Experiment result shows that this method not only effectively expands the semantic vocabularies of a 3D model library without abundant categories by enlarging the coverage of vocabularies, but also keeps good semantic relevance of the expanded vocabularies to the original ones, so as to achieves semantic based 3D model retrieval effectively.
出处
《计算机科学》
CSCD
北大核心
2009年第2期152-154,185,共4页
Computer Science
基金
国家自然科学基金优秀国家重点实验室专项基金资助项目(项目编号60723003)
国家自然科学基金资助项目(项目编号60505008)
江苏省自然科学基金创新人才(学术带头人)基金资助项目(项目编号BK2007520)资助
