Improvement of TF-IDF Weight Calculation Method in Text Classification

TF-IDF is a classical method for calculating feature weight calculation in text classification,but it does not consider the distribution of feature words in the document collection itself,which results in less classification.In this paper,the IDF function is improved by calculating the ratio of the intra-class density of the feature words to the overall average density of the feature words evenly distributed in the sample.Experiments show that the improved TF-IDF considers the intra-class distribution of feature words and the distribution of the overall document set,which improves the ability to distinguish categories and effectively improves the text classification effect.