【摘要】 针对现有数学检索排序方法所存在的排序属性单一问题,提出一种基于犹豫模糊距离的科技文档多属性综合排序方法。归纳科技文档的评价特征,确立了包括数学表达式几何属性、文档重要性属性和全局属性的三元组;建立以数学查询表达式为标准的科技文档犹豫模糊集,定义了各指标对应的犹豫模糊隶属度函数;利用犹豫模糊集在多属性决策方面的优势,采用文档相似度评价方法对检索结果进行排序。不同领域包含345 027个数学表达式的19 964篇科技文档的排序对比实势表明,此方法比科技文档单一属性排序方法的合理性有一定的提高。更多还原

【Abstract】 Aiming at the problem of the single sorting property of existing mathematical retrieval ranking methods,a multi-attribute comprehensive ranking method based on hesitant fuzzy distance was proposed. The evaluation features of scientific documents were summarized,a three tuple consisting of mathematical expressions and documents importance’s global attributes was established. A hesitant fuzzy set of scientific documents based on mathematical query expression was established,and the hesitant fuzzy membership function corresponding to each index was defined. Through employing the advantages of hesitant fuzzy sets in multiple attribute decision making,a ranking method of scientific documents was proposed through document similarity evaluation. The comparison experiments of 19 964 scientific and technological documents containing 345 027 mathematical expressions from different fields show that the ranking rationality of scientific documents improves the ranking rationality of scientific documents in comparison with single attribute ranking.更多还原