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

Authors : Seda POSTALCİOGLU, Senem AKTAS

Pages : 215-221

Doi:10.17350/HJSE19030000191

View : 21 | Download : 7

Publication Date : 2020-09-30

Article Type : Research Paper

Abstract :F or this study Sentiment Analysis SA is applied for the music comments using different Neural Network NN Models. SA is commonly used for Natural Language Processing NLP . With the help of NLP, the evaluations / tips about the future can be obtained by analyzing the correspondences and comments. The aim of the study is to draw conclusions from the comments made under the songs whether they are nostalgic. Data is captured using the YouTube Data API. Data extraction is done by entering the link of the song whose comments will be taken. CSV files are obtained and then labeled as nostalgic and non-nostalgic. Different neural network models as MLPNN Multi-Layer Perceptron Neural Network , CNN Convolutional Neural Network , RNN-LSTM Recurrent Neural Network-Long Short-Term Memory are applied for sentiment analysis. Their performances are analyzed. MLPNN, CNN, RNN-LSTM performance results are 78%,88%,88%, respectively
Keywords : CNN, MLPNN, NLP, RNN, YouTube comments