Modeling Dan Analisis Menggunakan Sentiment Analysis Mengenai Konten Youtube Pada Channel Gadgetin Dengan Metode Fine-Tune Menggunakan Pre-Trained Model

Abstract

Sentiment Analysis is the process of analyzing digital text to determine whether the emotional tone of a message is positive, negative, or neutral. The analysis is based on the results of a machine learning model's detection of comments data. The machine learning model is built using the fine-tuning method with a pre-trained model. The pre-trained model used is the Bidirectional Encoder Representations from Transformers (BERT), which is a popular model among researchers. BERT is a Natural Language Processing (NLP) model developed by Google. The aim of this research is to ascertain the Accuracy value of the BERT model for Sentiment Analysis on YouTube comments. The data is obtained from comments on a YouTube Channel's content, "GadgetIn," and is analyzed using the pre-trained BERT model.

The dataset used consists of 500 comments, where 400 comments are used as

Training data, 50 comments for validation, and 50 comments for Testing. The

results of this research show an Accuracy value of 80%.