CAJ | 학술논문

现有的科技奖励检索都是基于关键词的匹配，忽略了对用户查询的语义理解。针对传统信息检索存在的问题以及结合当前面向实际应用的基于本体的语义检索的特点，提出一种关联多本体的科技奖励检索方法。通过对科技奖励项目的领域本体构建，对概念相似度计算方法的改进及关联多本体概念，使扩展词语更能表达用户检索意图。实验结果表明，该方法对比传统查询检索和单独本体扩展检索获得了更好的召回率和准确率。
현유적과기장려검색도시기우관건사적필배，홀략료대용호사순적어의리해。침대전통신식검색존재적문제이급결합당전면향실제응용적기우본체적어의검색적특점，제출일충관련다본체적과기장려검색방법。통과대과기장려항목적영역본체구건，대개념상사도계산방법적개진급관련다본체개념，사확전사어경능표체용호검색의도。실험결과표명，해방법대비전통사순검색화단독본체확전검색획득료경호적소회솔화준학솔。
The existing science and technology award retrieval is based on keywords matching technology, which ignores the semantic understanding of user queries. To solve the problem of traditional information retrieval and combine characteristics of current practical application-oriented and ontology-based semantic retrieval, this pa-per proposes a science and technology award retrieval method of associating multiple-ontologies. By constructing the domain ontology for science and technology awards projects and by improving the concept similarity calcula-tion method and associating multiple-ontologies concepts, the proposed retrieval method makes the query expan-sion words further express retrieval intention of users. The experimental results show the method obtains better pre-cision and recall compared with keywords query and single-ontology extensions query.