Utilizing ChatGPT for Sentiment Analysis
Macháliková, Kristina (2023)
2023
All rights reserved. This publication is copyrighted. You may download, display and print it for Your own personal use. Commercial use is prohibited.
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2023121537710
https://urn.fi/URN:NBN:fi:amk-2023121537710
Tiivistelmä
The purpose of this study was to explore the conduct of sentiment analysis using ChatGPT. The emotions and opinions expressed by customers regarding their purchased services, goods or about the company, in general, provide valuable information for businesses, potentially influencing their decision-making processes.
In this thesis, various literary sources for its theoretical foundation, complemented by real-life examples, were utilized. Getting acquainted with the development of GPT models provided knowledge of their ever-improving ability to understand and generate human-like text. Next, an overview of aspect-based sentiment analysis, which identifies and evaluates sentiments tied to specific product or service aspects in the text, was presented. An essential aspect of this thesis focused on the necessity of conducting analysis with respect to data protection and privacy, which are critical for the reliability and ethical operation of a company.
From the sampled examples, it was shown that ChatGPT could read text directly from a screenshot and perform sentiment analysis as instructed. A more accurate analysis required ChatGPT to have access to the entire submission, including emoticons and hashtags.
In this thesis, various literary sources for its theoretical foundation, complemented by real-life examples, were utilized. Getting acquainted with the development of GPT models provided knowledge of their ever-improving ability to understand and generate human-like text. Next, an overview of aspect-based sentiment analysis, which identifies and evaluates sentiments tied to specific product or service aspects in the text, was presented. An essential aspect of this thesis focused on the necessity of conducting analysis with respect to data protection and privacy, which are critical for the reliability and ethical operation of a company.
From the sampled examples, it was shown that ChatGPT could read text directly from a screenshot and perform sentiment analysis as instructed. A more accurate analysis required ChatGPT to have access to the entire submission, including emoticons and hashtags.