A high-quality feature selection method based on frequent and correlated items for text classification

Abstract The feature selection problem is a significant challenge in pattern recognition, especially for classification tasks. quality of the selected features plays critical role building effective models, and poor-quality data can make this process more difficult. This work explores use association analysis mining to select meaningful features, addressing issue duplicated information features. A novel technique text proposed, based on frequent correlated items. method considers both relevance interactions, using as metric evaluate relationship between target was tested SMS spam collecting dataset from UCI machine learning repository compared with well-known methods. results showed that proposed effectively reduced redundant while achieving high accuracy (95.155%) only 6%