Sentiment Analysis of Twitter about ChatGPT
Author(s):
Article Type:
Research/Original Article (دارای رتبه معتبر)
Abstract:
In recent years, we have witnessed significant advancements in artificial intelligence across many aspects of human life. One way AI can enhance human life is through the use of chatbots. A chatbot that has recently been introduced with much attention and is promised to revolutionize the way people interact with technology is ChatGPT. However, with the widespread use of AI chatbots, concerns about data privacy and security have emerged. Evaluating these concerns can offer insights into public perceptions and help improve data privacy policies. Previous research on this technology has mainly focused on its technical aspects, whereas understanding public sentiment about ChatGPT as a transformative technology can provide insights into its potential success or failure, as well as its strengths and weaknesses. In line with this, the present study aims to examine the perceptions of Twitter users regarding ChatGPT through sentiment analysis and topic modeling. A total of 478,266 tweets were collected via the official Twitter API, and following sentiment analysis using the BERT model—one of the advanced algorithms in deep learning—the results showed an accuracy of 82%. Additionally, through topic modeling using the BERTopic algorithm, based on BERT, the results achieved a coherence (C_V) score of 0.632 and a U_Mass score of -2.957. According to the study’s findings, the nine most discussed topics among Twitter users are: artificial intelligence, search engines, future jobs, answering questions, education, programming, large language models, business, and healthcare. The results indicate that users expressed the highest percentage of positive sentiment towards the topics of large language models, education, and business, while the most negative sentiments were expressed regarding future jobs, healthcare, and artificial intelligence. After neutral opinions, which made up the largest portion of the data, positive tweets significantly outnumbered negative ones, reflecting the public’s satisfaction and optimism towards ChatGPT technology.
Keywords:
Language:
Persian
Published:
Information management, Volume:9 Issue: 1, 2023
Pages:
179 to 202
https://www.magiran.com/p2847665
سامانه نویسندگان
مقالات دیگری از این نویسنده (گان)
-
Solving the Traveling Salesman Problem Using a Modified Teaching-Learning Based Optimization Algorithm
Ahmad Aliyari Boroujeni, *
International Journal of Industrial Engineering and Productional Research, Jun 2025 -
Utilizing Deep Learning for Aspect-Based Sentiment Analysis in Restaurant Reviews
*, Fatemeh Abbasi, Seyedeh Armina Mohseni, Raha Basraei
Journal of Information Technology Management, Spring 2025