Predicting Emotion in Movie Scripts Using Deep Learning
Dong-Min Kim, Seong-Ho Lee, Yun-Gyung Cheong · 2018
In this paper we present a machine learning approach that aims to predict an emotion that follows given text segments. For this, we pre-processed a collection of movie scripts to generate a sequence of sentiment values. The sequence is given to a deep learning network to learn the sentiment value to follow. The evaluation result was promising, showing high cosine similarities between the target and the predicted emotion values.