Jurnal Locus Penelitian dan Pengabdian
Vol. 4 No. 9 (2025): : JURNAL LOCUS: Penelitian dan Pengabdian

Sentiment Analysis Of Reviews On The Chatgpt Application Using Long Shortterm Memory Method

Pramudita, Dias Aziz (Unknown)
Imaduddin, Helmi (Unknown)
Afiana Azizah, Synta Nur (Unknown)
Muslihah , Isnawati (Unknown)



Article Info

Publish Date
11 Sep 2025

Abstract

The use of information and communication technology (ICT) in the era of Education 5.0 has transformed how humans interact and learn. A significant development in this context is the application of artificial intelligence (AI) in the education sector, which has provided substantial benefits while also posing challenges, including data privacy and security issues. Chatbots using the GPT (Generative Pretrained Transformer) model are a prominent example of AI application. This study aims to analyze user sentiment towards the ChatGPT application on the Google Play Store using the Long Short-Term Memory (LSTM) method. The analysis encompasses reviews ranging from a rating of 1 (negative) to a rating of 5 (positive) out of a total of 193,749 reviews with an average rating of 4.9. Final data shows that 8,927 comments were collected for initial analysis, followed by preprocessing and analysis. Testing results indicate that the LSTM method achieved an accuracy of 76,80%, precision of 71%, recall of 70.%, and an F-measure of 76%. It can be concluded that the chatgpt application can function well and be received positively, which can then integrate this technology into the educational curriculum wisely, without replacing the important role of teachers in the learning process.

Copyrights © 2025






Journal Info

Abbrev

jl

Publisher

Subject

Aerospace Engineering Automotive Engineering Decision Sciences, Operations Research & Management Electrical & Electronics Engineering Mathematics

Description

Jurnal Locus: Jurnal Ilmiah Penelitian dan Pengabdian, double-blind and open-access academic journal in the Multidisiplin. This journal is published once a month by CV. Riviera ...