使用最大熵模型进行文本分类

最大熵模型是一种在广泛应用于自然语言处理中的概率估计方法。文中使用最大熵模型进行了文本分类的研究。通过实验,将其和Bayes、KNN、SVM三种典型的文本分类器进行了比较,并且考虑了不同特征数目和平滑技术对基于最大熵模型的文本分类器的影响。结果显示它的分类性能胜于Bayes方法,与KNN和SVM方法相当,表明这是一种非常有前途的文本分类方法。

Using Maximum Entropy Model for Text Categorization

Maximum Entropy Model is a probability estimation technique widely used for a variety of natural language tasks.This paper uses maximum entropy model for text categorization.It compares and analyzes its categorization performance using different approaches for different number of feature and smoothing technique.Moreover,in experiments it compares it to Bayes,KNN and SVM,and shows that its performance is higher than Bayes and comparable with KNN and SVM.The author thinks it is a promising technique for text categorization.