Comparison of Neural Network Models for Nostalgic Sentiment Analysis of YouTube Comments

Abstract

W ith the rapid development of social networking sites, people need to explain their opinions and thoughts. People share not only information with websites but also they begin to express their thoughts and feelings more easily [1]. They expressed these posts by tweeting them on Twitter, posting on Instagram or commenting on social media. Based on this case, the feelings and thoughts of the interpreters can be analyzed [1]. Reference 11 can be shown as an example. It is possible to analyze such comments or posts according to some criteria such as what team hits, political opinion, positive/negative/neutral content of comments. This kind of analyzes may affect future production progress by enabling the public to learn the opinions, parties, and thoughts before the political election or to make a customer analysis for the products produced. It has a wide usage area. These analyses are carried out for many different purposes. With these studies, it is tried to predict the future. These analyses are referred to as "Sentiment Analysis" in terminology [1]. Sentiment analysis is one of the techniques commonly used in Natural Language Processing (NLP). NLP helps us to understand the content of a text [2]. There are different