Code

This compressed package contains the ATF-DF-VS algorithm code folder, TF-IDF-KNN algorithm code folder, in which the README file in the VS and KNN folders is the code description.Here's a description of the study:In this study, we propose the ATF-DF-VS algorithm for text classification. The algorithm aims to extract class-typicalfeature vectors from labeled text samples, andbased on these extractedvectors, the classification process by leveraging vector sparsity isfinally completed. The algorithm is structured into two distinct stages. The first stage is the typical feature extraction stage. In this part, an ATF-DF algorithm is introduced to extract the distinctive features of large text datasets with predefined categories. The ATF-DF weight values for all terms in the characteristic dictionary are calculated. Following a DF descending order and data compression procedure, the feature terms with the highest values are selected as the class-typicalfeature vectors. The second stage is the text classification stage. The classification samples are vectorized using the class typical feature vectors, and several vectors of the samples to be classified are obtained. Text classification is completedby calculating the sparsity of the vectors. This algorithm leverages the advantages of big data sample statistics and has a unique advantage in text classification calculations. The experimental findings indicate that the ATF-DF-VS algorithm outperforms traditional text classification algorithms in accurately extracting class- typicalfeature vectors from various types of text. This enhancement leads to a marked improvement in text classification accuracy, a substantial reduction in computational requirements for text classification, and an increase in processing speed.