摘要： 针对经典粗糙集模型难以分类标引空间以及体现类间关联的缺陷，将条件概率关系结合粗糙集理论引入信息检索，提出一种基于概率粗糙集的信息检索模型。定义标引词空间的条件概率关系，自动挖掘概念相似类形成概念空间。定义文档与查询、文档与文档间语义贴近度的计算方法。根据贴近度实现检索匹配结果的排序输出。仿真实例表明了该方法的可行性和有效性。

关键词: 粗糙集, 信息检索, 条件概率关系, 语义贴近度

Abstract: Aiming at the disadvantage of classical rough set theory on identifying the conceptually similar terms and the relationships between classes, this paper proposes a novel information retrieval model based on conditional probability relation and rough set. Conception space is formed by defining conditional probability relation in index words space to mine conception similar class automatically. A method is designed to calculate semantic distance between a document and a query, as well as documents. And the ordered outputs of retrieval result are acquired. The simulation instance shows that this algorithm is feasible and effective in practice.

Key words: rough set, information retrieval, conditional probability relation, semantic distance

中图分类号: 

• TP391