Efficient classification model of web news documents using machine learning algorithms for accurate information. Issue 98 (November 2020)

• Record Type: Journal Article
• Authors: Mulahuwaish, Aos; Gyorick, Kevin; Ghafoor, Kayhan Zrar; Maghdid, Halgurd S.; Rawat, Danda B.
• Is Part Of: Computers & security. Issue 98(2020)
• Publication Date: 2020-11
• Subjects: Web News Applications -- Data mining; SVM -- kNN -- DT -- LSTM; Computer security -- Periodicals; Electronic data processing departments -- Security measures -- Periodicals; 005.805
• DOI: 10.1016/j.cose.2020.102006
• Languages: English
• ISSNs: 0167-4048

Abstract

Abstract: Web applications are regarded as a popular platform to exchange information with users. These applications have to be able to process Big-Data quickly and to serve users in a timely manner with accurate information posted in news portals which can be a huge challenge to overcome. Huge computation power is needed to crawl the web and process big-data and the methods are needed to be developed to reduce space and time complexity of this process. Data mining is considered to be a solution to mitigate the aforementioned challenges by extracting specific information based on explicit features. This paper proposes an efficient model for web that extracts news information and sorts news documents into four different categories business, technology & science, health and entertainment. Four different machine learning classifiers Support Vector Machine (SVM), K-Nearest Neighbors (kNN), Decision Tree (DT) and Long Short-Term Memory (LSTM) are compared. These classifiers are implemented separately and are then compared using accuracy and receiver operating characteristic curves. The attained results show that the accuracy of kNN was the worst at 88.72% and SVM was the best at 95.04%.