基于Markov模型的改进型疑错窗口算法

提出了一种改进型疑错窗口方法,同时基于该方法提出了一种新的中文文本校对方法。该方法预先用窗口技术初步定位可能存在错误的语段,然后再在各个疑错窗口内运用Markov模型进行定错和校对。由于该新方法较大程度地减小了校对候选矩阵的维数,因而相比传统的Markov模型,该方法有更小的计算复杂度。试验结果表明,相比传统方法,新方法在时间复杂性和空间复杂性上得到了明显降低,而纠错效果基本不变。

Algorithm for Improved Window Technology Based on Markov Model

This paper first presents an improved window technology.Then a new automatic Chinese text proofreading method is proposed.This new approach consists of two stages:firstly,the improved window technology algorithm is used to locate error text segment.The second step is to use Markov model to detect and correct such error in each window.Results indicate that our algorithm has similar performance with standard Markov model method.But in contrast with standard Markov model method,the new algorithm has lower complexity in space and time.