中文摘要：

【目的／意义】面对海量非结构化的数据，如何快速准确地检索到目标信息，实现相关信息的自动关联，是语义检索和智能推荐的研究重点。【方法／过程】为解决该问题，提出了一种基于本体的JESS推理模型。以目前常用的形式概念分析法（FCA）（应用于知识采集）及描述逻辑（应用于知识表达）为基础进行相关术语和概念的抽取，探讨两者在使用过程中的问题及不协调处，并提出改进方式。在基于LCS原则上，探索新增概念及上下级关联架构。然后利用本体建模工具protege构建领域本体，建立基于推理引擎JESS的检索查询系统，进行检索。【结果／结论】实证研究表明，该本体模型支持基于语义推理的智能查询，并能提高查全率及查准率。

英文摘要：

[ Purpose/significance] In the environment of massive unstructured data, it is very important for semantic retriev- al and intelligent recommendation to quickly and accurately retrieve the target information and realize the automatic correla- tion of relevant information. [Method/process] This study proposes a JESS reasoning model to solve the problem. Using FCA and DL extracts conceptual structures and conceptual contents. Exploring the new concept and conceptual structures based on LCS. Then using protege to build domain ontology and creating JESS reasoning exploring system. [ Result/conclu- sion]The result shows that the model supporting intelligent recommendation based on the semantic reasoning with a high ef- ficiency.