Treffer: Developing an Application for Articles Classification Using the KNN Algorithm.

Classifying textual data is essential in the rapidly evolving field of computational linguistics. The K-nearest neighbors (KNN) algorithm, a supervised learning algorithm, is popular for regression and classification due to its simplicity and effectiveness. Previous studies have demonstrated the usefulness of KNN across various areas, including image recognition, recommendation systems, and text classification. However, its role in classifying articles, particularly with N-grams and distance metrics, has not been thoroughly explored. This research intends to fill these gaps by creating a Java application that implements the KNN algorithm to classify articles based on their content. The application utilizes trigrams for text comparison and assesses the performance of three distance metrics: Euclidean, Manhattan, and Chebyshev. By analyzing the effects of these metrics alongside various feature selection techniques, this study aims to enhance the accuracy and efficiency of article classification. [ABSTRACT FROM AUTHOR]