摘要： 传统的案例检索算法较少考虑多影响因子对属性权值的作用及属性间关联性。为此，结合改进覆盖算法和灰色关联度优化案例检索，提出基于多影响因子的案例属性权值自适应算法和改进灰色关联度的案例相似度算法，解决案例中属性权值确定和相似度计算问题。实验结果表明，与多种案例检索算法相比，该算法能够克服传统案例检索的缺陷，有效提高案例检索的质量和效率。

关键词: 案例检索, 相似度计算, 覆盖算法, 灰色关联度, 特征权值自适应算法

Abstract: Multi-influence factor effect on the attribute weight and the relatedness of its attributes is rarely considered in traditional case searching algorithm. Therefore, this paper proposes an improved case similarity algorithm combined the improved covering algorithm and the grey relational degree to optimize case retrieval. Proposed algorithm obtains the best retrieval case by using the improved grey relational degree case retrieval algorithm whose case attribute weight is figured out by auto-adapted algorithm based on the multi-influence factor. Compared with many other kinds of case searching algorithms, experimental results indicate that proposed algorithm can overcome the disadvantage of the traditional case retrieval algorithm and enhance the quality and the efficiency of the case retrieval effectively.

Key words: case retrieval, similarity computation, covering algorithm, grey relational degree, characteristic weight auto-adapted algorithm

中图分类号: 

• TP182