【摘要】 提出一种基于向量空间模型的科技文献自动分类方法。首先构建分类代表文本,计算单文本与分类代表文本之间的相似度;然后利用训练集计算出分类归属度,以判断某一文本分类的可能性,实现科技文献的自动分类。以3个学科的文献为例,验证该方法的可用性。更多还原

【Abstract】 The paper gives an automatic classification method of sci-tech document basing on vector space model. Firstly it constructs classification representative text and computes degree of similarity between single text and classification representative text; then it uses training set to calculate membership degree of classification, so as to estimate possibility of text classification and realize automatic classification of sci-tech document. The paper takes document of three disciplines for example to verify usability of this method.更多还原