期刊名称：Journal of Theoretical and Applied Information Technology
印刷版ISSN：1992-8645
电子版ISSN：1817-3195
出版年度：2021
卷号：99
期号：8
语种：English
出版社：Journal of Theoretical and Applied
摘要：Fake news is a problem that has a major effect on our life. Detection of fake news considered an interesting research area that has some limitation of the available resources. In this research, we propose a classification model that is capable of detecting fake news based on both Doc2vec and Word2vec embedding as feature extraction methods. Firstly, we compare between the two approaches using different classification algorithms. According to the applied experiments, the classification based on Doc2vec model provided promising results with more than one classifier. The Support vector machine resulted the best accuracy with 95.5% followed by Logistic Regression 94.7% and the Long Short Term Memory produced the lowest accuracy. On the other hand, the classification based Word2vec embedding model results high accuracy only with Long Short Term Memory classifier with 94.3%. Secondly, the classification models based on proposed Doc2vec have shown to outperform a corresponding model that based on TF-IDF on the same dataset using Support Vector Machine and Logistic Regression classifiers.
