KLASIFIKASI TEKS BUATAN AI (ARTIFICIAL INTELLIGENCE) MENGGUNAKAN METODE BIDIRECTIONAL LONG SHORT TERM MEMORY (BILSTM)

ALAMSYAH, M.RENDI and Utami, Alvi Syahrini (2024) KLASIFIKASI TEKS BUATAN AI (ARTIFICIAL INTELLIGENCE) MENGGUNAKAN METODE BIDIRECTIONAL LONG SHORT TERM MEMORY (BILSTM). Undergraduate thesis, Sriwijaya University.

	Text RAMA_55201_09021282126061.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (3MB) | Request a copy
	Text RAMA_55201_09021282126061_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7MB) | Request a copy
	Text RAMA_55201_09021282126061_0022127804_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB)
	Text RAMA_55201_09021282126061_0022127804_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (516kB) | Request a copy
	Text RAMA_55201_09021282126061_0022127804_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (442kB) | Request a copy
	Text RAMA_55201_09021282126061_0022127804_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (656kB) | Request a copy
	Text RAMA_55201_09021282126061_0022127804_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (495kB) | Request a copy
	Text RAMA_55201_09021282126061_0022127804_06.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (188kB) | Request a copy
	Text RAMA_55201_09021282126061_0022127804_07_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (199kB) | Request a copy
	Text RAMA_55201_09021282126061_0022127804_08_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (285kB) | Request a copy

Abstract

The rapid development of Artificial Intelligence (AI) enables this technology to be used for various purposes, one of which is generating text that resembles human writing. This capability brings a new challenge, especially in distinguishing between text generated by AI and text created directly by humans. This study utilizes the Bidirectional Long Short-Term Memory (BiLSTM) method and Word2Vec word embedding to classify AI-generated text as a solution to this challenge. The data used consists of 23,000 essay texts divided into two classes, human and AI. The results show that the best BiLSTM model, with a hyperparameter configuration of learning rate 0.0001, 128 hidden units, 0.2 dropout, batch size of 32, and 15 epochs, achieves excellent performance in classifying AI-generated text, with an accuracy of 99.30%, precision of 99.56%, recall of 99.04%, and an f1-score of 99.30%.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Artificial Intelligence, Bidirectional Long Short-Term Memory, Klasifikasi Teks, Word2Vec
Subjects:	T Technology > T Technology (General) > T1-995 Technology (General)
Divisions:	09-Faculty of Computer Science > 55201-Informatics (S1)
Depositing User:	M.Rendi Alamsyah
Date Deposited:	07 Jan 2025 02:06
Last Modified:	07 Jan 2025 02:06
URI:	http://repository.unsri.ac.id/id/eprint/162761

Actions (login required)

View Item