Distribution of Polarity Value between VADER and TextBlob in Sentiment Classification of Tourist Vlog Content Reviews

Singgalen, Yerik Afrianto

JOURNAL OF INFORMATION SYSTEM RESEARCH (JOSH)
Vol 5 No 4 (2024): Juli 2024

This research employs sentiment analysis techniques to examine audience perceptions across three videos featuring tourist vlog content. Utilizing the CRISP-DM framework, the study compares the performance of VADER and TextBlob in sentiment classification, analyzing the distribution of polarity values and agreement levels between the two models. The findings reveal varying proportions of negative, neutral, and positive sentiments across the videos, with VADER and TextBlob demonstrating fair agreement levels ranging from 64.97% to 72.60%. These results underscore the importance of employing diverse sentiment analysis tools and language-specific models for accurate sentiment classification. The research contributes valuable insights for content creators, marketers, and analysts in understanding audience sentiments and shaping content strategies effectively.