Scamming Detection from social media using Deep Learning Approach

Amare, Lakew

dc.contributor.author	Amare, Lakew
dc.date.accessioned	2024-04-09T05:55:57Z
dc.date.available	2024-04-09T05:55:57Z
dc.date.issued	2023-02
dc.identifier.uri	http://ir.bdu.edu.et/handle/123456789/15759
dc.description.abstract	Scamming is the practice of using tricks to obtain money or goods from another person. Scams often involve making false promises of quick riches or easy money or tricking victims into handing over personal or banking information. Scamming detection technology is constantly evolving and improving, but it remains a challenging task due to the imagination and creativity of scammers. Scamming detection on social media is a critical issue that needs to be addressed to secure online transactions and protect individuals from financial fraud. Scams can be difficult to identify since they come in many forms and are constantly changing. The increasing usage of social media platforms has led to a rise in the number of scams carried out through these platforms. To tackle this problem, we used deep learning approaches to detect scammers in real-time on social media. Scamming distribution has become a significant issue for societies worldwide due to the use of such convenient technology for immoral purposes. This phenomenon is especially common during online business transactions. In this study, we used the collected data from a free dataset repository and an annotated dataset of around 6549 and above texts to build an automated scamming detection system using deep learning algorithms. We used an experimental research methodology to develop a model for scamming detection in this research work. We have used recurrent neural networks (RNN) and convolutional neural networks (CNN) to design a model. We perform several experiments to determine the best-performing deep learning algorithms. The CNN model is better than the RNN model, with an accuracy of 91% and an RNN of 85%. In addition to that, we compared the CNN and RNN algorithms and tested them. Future research directions in the problem domain include identifying the scammers. And, to improve the model, increase the number of datasets used to train the model. In addition to our contribution to the analysis of these deep learning models, we believe that the proposed model will create an environment for more studies and findings concerning the detection and prevention of scams by collecting Amharic dataset. Keywords: social media, scamming, Deep learning, Natural Language Processing	en_US
dc.language.iso	en_US	en_US
dc.subject	Software Engineering	en_US
dc.title	Scamming Detection from social media using Deep Learning Approach	en_US
dc.type	Thesis	en_US