Effectiveness of Feature Selection in Text Summarization

Text summarization and classification are crucial techniques for managing the massive and steadily growing volume of text documents published every day. Feature Selection (FS) is a significant problem in text summarization and classification. This paper provides a summary of the most common feature selection methods and recent techniques depending on feature selections utilized in text summarization and classification. It includes a discussion of 13 scientific publications. The feature selection methods discussed in the paper are unsupervised such as correlation and supervised such as embedded, filter, and wrapper. The feature selection techniques that are presented in this paper are correlation, Term Frequency-Inverse Document Frequency(TF-IDF), textrank, x^2-statistic (CHI-square), mutual information, itemset mining, Information Gain (IG), Improved Fruit Fly Optimization Algorithm (IFFOA), Firefly Algorithm (FA), Harris Hawks Optimization (HHO), and Support Vector Machine-Recursive Feature Elimination(SVM-RFE).