摘要： 根据改进的词语权重计算方法构造给定文本的特征向量,并用之从专利分类的领域知识――国际分类表IPC中直接提取类别的概念向量和待分类专利文本的特征向量,然后采用向量空间模型实现专利的自动分类,该方法不需要大量的训练样本,具有较高的分类正确率和执行速度。

关键词: 专利分类, 向量空间模型, 领域知识

Abstract: This paper improves the classical approach for calculating the term weight in text,and then employs it to construct the concept vector of the class directly from the domain knowledge of the international patent classification(IPC) and the feature vector of the patent text to be classified.Subsequently vector space model is utilized to classify the given patent text.The experiment and practical systems indicate that the approach is of high classification precision and efficiency.

中图分类号: 

• TP391.1