基于BP神经网络的中文分词算法研究

Approaching Chinese Words Segmentation Algorithm Based on BP Neural Network

探讨了BP神经网络的学习过程与主要参数,分析了基于BP神经网络的中文分词算法,在joone-editor建立好的神经网络模型中加以实验,实验表明面对歧义切分的时候,BP神经网络的学习和适应能力较原先的MMAnalyzer具有明显的进步。

This paper briefly discusses the learning process and main parameters of BP neural network,and then analyses implementation of Chinese words segmentation based on BP neural network,and finally experiments on the neural network model establishing by joone-editor.It concludes that the learning and adaptability of BP neural network is more efficient than the original MMAnalyzer in ambiguous segmentation.