【摘要】 文章在表示文献向量的关键词向量组相关的假设下,提出了非正交基的向量空间检索模型的概念,并导出计算关键词对相关性的相关矩阵和相应匹配函数的算法。这可作为索引关键词的自动分类和改善各种检索策略中关键词相关性的基础。更多还原

【Abstract】 On the assumption that there is correlation among term vectors for rep-resenting document vectors, this paper proposes the concept of non-orthogo-nal vector space model for representing document vectors, thereby, deducingthe correlation matrix for calculating the term-pair correlationship and thecorresponding matching function algorithm which can be used for the automa-tic classification of indexing terms and serves as the basis for improving theindex term correlationship among various strategies.更多还原