Analysis of Sentence Ordering Based on Support Vector Machine

In this paper, we propose a practical approach to sentence ordering in multi-document summarization taks of Chinese language. First, by using the support vector machine (SVM), we classify the sentences of a summary into several groups of rough position according to the information of the source documents. After classification, we adjust the sentence sequence of each group with information of feature-adjacency learned from raw corpus, and find the sequence of each group. Then we connect the sequences in different groups to generate the final order of the summary. Experimental results indicate that this method works better than most existing methods of sentence ordering.