Exploring Chinese Dynamic Sentiment/Emotion Analysis with Text Mining—Taiwanese Popular Movie Reviews Comment as a Case

Abstract

With the rapid development of technology, viewers’ opinions are no longer passed on by word-of-mouth in the real world but expressed on the Internet. In the six categories of life, food, clothing, housing, travel, education, and entertainment, every part is inseparable from comments. The most common entertainment in the six majors is watching movie. When people are interested in a movie, they might collect information online firstly so they might spend a lot of time reading online reviews and ratings. In order to have a general understanding of the movie, they will check different reviews and ratings on different online platforms. Thus, it is very meaningful to sort out helpful reviews from a large number of reviews and comments from other people. In this paper, we apply text exploration technology to the comments of movie and the analysis will allow the viewers to clearly identify the information they need. This study collects movie reviews from the online platform (YAHOO film) for text exploration and analysis. Many comments don’t align with the ratings, so the primary goal of research is to verify whether the netizens’ comments are consistent with the ratings, then discuss the connection between movie date and the emotional words. The viewers can refer to these overall analysis and emotional time dynamics in the future when they want to check feedbacks from other people. The result shows that the time dynamic analysis has significant differences in emotional words in the third week so it would be better to the viewers referring to the third-week rating as a quick check before watching movies.