基于中文分词的文本相似度动态规划算法  

Dynamic Programming Algorithm of Text Similarity Calculation Based on Chinese Word Segmentation

作　　者：肖侃[1] 谭长庚[1] 丁玲[1] 

机构地区：[1]中南大学软件学院,湖南长沙410075

出　　处：《现代电子技术》2011年第8期72-74,78,共4页Modern Electronics Technique

摘　　要：针对传统的基于动态规划的对论文的文本相似度计算的不足,提出了一种基于中文分词和动态规划的论文文本相似度计算方法,并对此进行了讨论。该方法克服了一般基于动态规划的计算方法所具有的效率低,判断准确率低的缺点。通过对实际中的论文数据库中论文进行测试和分析,该算法能提高计算准确率,并一定程度上提高了运算速度,可以应用于论文防抄袭系统中。A dynamic programming algorithm of similarity calculation of thesis texts based on Chinese word segmentation and dynamic programming are proposed and discussed to overcome the deficiency of those traditional algorithms.This algorithm solved the problems of inefficiency and low preciseness existing in traditional dynamic algorithms.The testing and analysis for the theses in a real thesis database prove that the algorithm can improve the calculation accuracy and the analysis speed in some degree,and can be applied to the Thesis Anti-plagiarism System.

关 键 词：文本相似度 动态规划 最长公共子序列 中文分词 

分 类 号：TN919-34[电子电信—通信与信息系统]