Sentiment Analysis on Adventure Movie Scripts

摘要

Sentiment analysis is the process of mining subjective information in different kinds of sources for playing an important role in making decisions. Movie scripts are an interesting text-domain including their diverse expressions for sentiment mining. In this paper, we have done sentiment analysis by applying Minimum Viable Product (MVP), Exploratory Data Analysis (EDA), and TextBlob sentiment analysis techniques. We compute the sentiment polarity and show the fluctuation of sentiments over time of a specified movie genre. We collected a dataset consisting of 20 adventure genre’s movie scripts for sentiment analysis on movie scripts. The experimental results show that the fluctuations of the sentiment of the adventure genre’s movie reflect the viewer’s mindset as public opinion for social impact.