中文摘要：

现有的文献聚类方法都是通过文献关键词来进行的。本文在研究大量文献聚类方法的基础上,提出了一种通过文献关键词和摘要进行加权的新的文献聚类算法。首先,改进了传统相似度计算的方法,设计出基于关键词和摘要词加权的相似度公式,使文献相似度计算更加精确。其次,基于＂文献距离越大,聚为一类的概率越小＂的思想,提出了一种＂最大距离聚类法＂,并给出了算法的详细步骤。最后,实现算法并进行了大量的实验仿真。通过改进相似度计算公式,调整关键词和摘要词的权重,提高了聚类的质量。结果表明,本文提出的文献聚类算法是一种行之有效的方法。

英文摘要：

All document clustering methods are based on keywords now.By researching a lot of methods for document clustering,a new dynamic method is presented,which is based on the idea that the longer is the distance between two documents, the lesser probability that they can be classified in same class.Documents＇ similar matrix is computed accurately by a new formula based on keywords and abstract correlation.The steps of the algorithm are given in detail.Experimental results show that this is an effective method and the quality of clustering is improved by combining keywords with abstract＇ s weight.