【摘要】 现有的文本分类方法在知识获取方面存在不足。该文针对某种应用需求,提出了人工神经网络和文本分类结合的一种文本分类方法。采用特征词的向量空间来描述文本,利用人工神经网络的良好的学习能力,通过对文本样本集进行训练,从中提取出对文本分类的知识,再利用神经网络和所获得的分类知识实现对文本的分类。更多还原

【Abstract】 Existing methods of text categorization(TC)are defective to some extent in knowledge retrieval.Here we render a new method of TC using artificial neural networks(ANN)in response to the requirement of some domains.By describing documents with vector space of key words and training a″training″set of preclassified documents,we can retrive the required knowledge of categorization with the help of ANN’s perfect learning ability.At last we can apply the knowledge in TC.更多还原