Loading [a11y]/accessibility-menu.js
Classification ChatGPT Generated News and True News Using Support Vector Machines | IEEE Conference Publication | IEEE Xplore

Classification ChatGPT Generated News and True News Using Support Vector Machines


Abstract:

This study intends to demonstrate the news provided by ChatGPT and real news, as well as to show ChatGPT's capability to provide regular news. In this study, we used Supp...Show More

Abstract:

This study intends to demonstrate the news provided by ChatGPT and real news, as well as to show ChatGPT's capability to provide regular news. In this study, we used Support Vector Machine (SVM) to distinguish between ChatGPT-generated and real-world news. SVM is a powerful tool for data classification. Previous SVM research did not classify the news generated by ChatGPT. Since its inception in late 2022, ChatGPT has been amazing because it has become popular, and many researchers have been discussing it in numerous fields. In particular, ChatGPT-generated news refers to news items or other information produced or developed by ChatGPT using an AI language model. To extract the dataset and determine the meaning of the data, we employed Neural Language Processing (NLP) and a TDM Matrix. After that, we used Support Vector Machines (SVM) to categorize both the real and ChatGPT-generated news. The model's percentage of accurate predictions in both the positive and negative classifications is shown by the total accuracy value of 0.9848. It suggests that the model's accuracy rate is approximately 98.48%.
Date of Conference: 09-11 November 2023
Date Added to IEEE Xplore: 21 December 2023
ISBN Information:

ISSN Information:

Conference Location: Taichung, Taiwan

Funding Agency:


References

References is not available for this document.