Exploring the Influencing Factors of IP Film Rating by Sentiment Analysis and GMM

Abstract

Recently, intellectual property (IP) film has become an important accessory for entertainment, and its rating has become the focus of quality evaluation. However, existing research seldom conducts study on influencing factors of rating. In this paper, we use sentiment analysis and generalized method of moments (GMM) to explore the factors that affect IP film rating. We take advantage of production, broadcast, genre and audience feedback to construct six explanatory variables, including actor influence, screenwriter participation, broadcast time, broadcast platform, genre, and adaptation satisfaction. We use LLC, IPS and Sargan tests to conduct variable stability test and model setting test. From the regression results of 134 IP films that obtained by sample filtering, the impact of each influencing factor on the rating is obtained. We found that short-term historical rating, actor influence, adaptation satisfaction and screenwriter participation positively affect current rating. While, long-term historical rating has a negative impact on current rating. In addition, broadcast time and broadcast platform have imposed positive impact on IP film rating, and genre has only a weak impact on rating. Our work provides advice for IP film producers, prompting them to improve quality by emphasizing celebrity effects and author participation.