Analisis Sentimen Masyarakat Terhadap Generasi Z Dalam Dunia Kerja Pada Media Sosial Twitter Menggunakan Metode Naïve Bayes
Total View This Week0
Total View This Week0
Institusion
Universitas Muhammadiyah Malang
Author
Herlinda, Vita Amalia
Subject
Q Science (General)
Datestamp
2024-07-31 08:06:05
Abstract :
Generation Z, born between 1996 and 2010, is now entering the workforce and is known as "Digital Natives" due to their familiarity with technology and social media such as Twitter. They are actively sharing thoughts and experiences through Twitter, which has become the primary platform for open discussion and sharing perspectives. This research employs the Naïve Bayes method to analyze sentiment in tweets related to Generation Z in the workplace, utilizing the Inset lexicon labeling.The results indicate that the Inset lexicon labeling achieved an accuracy of 76%. Furthermore, experiments with manual labeling yielded an accuracy of 87%, demonstrating a significant difference compared to the lexicon?based approach. The study also examined the influence of stopwords during preprocessing, revealing that the lexicon-based labeling without stopwords achieved 96% accuracy, while manual labeling without stopwords reached 90%.These findings highlight the high sensitivity of lexicon-based models to the use of stopwords and their potential to enhance term variation in sentiment analysis of Generation Z in the workplace via Twitter.
Generation Z, born between 1996 and 2010, is now entering the workforce and is known as "Digital Natives" due to their familiarity with technology and social media such as Twitter. They are actively sharing thoughts and experiences through Twitter, which has become the primary platform for open discussion and sharing perspectives. This research employs the Naïve Bayes method to analyze sentiment in tweets related to Generation Z in the workplace, utilizing the Inset lexicon labeling.The results indicate that the Inset lexicon labeling achieved an accuracy of 76%. Furthermore, experiments with manual labeling yielded an accuracy of 87%, demonstrating a significant difference compared to the lexicon?based approach. The study also examined the influence of stopwords during preprocessing, revealing that the lexicon-based labeling without stopwords achieved 96% accuracy, while manual labeling without stopwords reached 90%.These findings highlight the high sensitivity of lexicon-based models to the use of stopwords and their potential to enhance term variation in sentiment analysis of Generation Z in the workplace via Twitter.
Download
book
BibTex
Latex, Jabref
cloud_download
Original Resource
url resource Institution
