【作者】 贾雪峰；

【导师】 王建新；

【作者基本信息】 北京林业大学 ， 计算机软件与理论， 2010， 硕士

【摘要】 目前的信息检索工具主要是基于关键词匹配或内容分类目录进行查找,很少做进一步的智能化处理。因此,反馈给用户的是简单的相关信息,无法满足用户的检索要求。本体论是对信息概念化的精确描述,利用本体对资源的描述来检索,可以一定程度上提高检索的效率和精确度。因此,基于本体的信息检索系统成为研究的重点。本文首先在研究主题词表和领域本体相关知识及构建过程的基础上,以木材科学与工程为例进行了林业领域本体的试构建,然后在对比传统检索模型的基础上,结合本体概念,提出了一种基于领域本体的智能检索模型。该模型利用本体中语义概念关系及语义扩展机制对查询关键字进行概念映射及扩展,返回检索结果,通过原型系统进行实验验证,证明该模型是对传统检索模型的改进,一定程度上改善了检索的查准率和查全率。最后,论文对研究中存在的问题进行了总结,并对今后的工作进行了展望。更多还原

【Abstract】 Nowadays, many information retrieval tools search information based on keyword matching or catalogue matching, and rarely process more intelligently. Consequently, they get back to users with very simple and limited information,which may not be very relevant and barely meet user’s expectation. Ontology is an explicit specification of a conceptualization. Information retrieval systems can be more effective and accurate by applying ontology to depict resources. Hence, the information retrieval system based on ontology becomes the main focus.Firstly, the author studies the thesaurus, domain ontology construction process, and other relevant knowledge, and then builds the wood science and engineering ontology as a trial. Comparing with the traditional retrieval model, a retrieval model based on domain ontology integrating with ontology concepts is proposed. Secondly, the prototype information retrieval system is established using the semantic relation and semantic extension, which expands the querying condition. Via the prototype model experiment data validation, it is proved that both precision and recall ratio are improved in a certain extent. At last, the existing problems are summarized, and the future researches are prospected.更多还原