Please use this identifier to cite or link to this item:
http://repository.ipb.ac.id/handle/123456789/153582
Title: | Analisis Sentimen Komentar Trailer Film Oppenheimer Pada YouTube Menggunakan GRU dan FastText Embedding |
Other Titles: | Sentiment Analysis of Oppenheimer Movie Trailer Comments on YouTube using GRU and FastText Embedding |
Authors: | Ardiansyah, Firman Haryanto, Toto Ansari, Aysuka |
Issue Date: | 2024 |
Publisher: | IPB University |
Abstract: | Pada era internet saat ini, trailer umumnya diunggah melalui platform media
sosial seperti YouTube. Melalui platform ini, penonton dapat mengakses serta
memberikan sentimen mereka terhadap trailer film dengan mudah. Analisis
sentimen terhadap komentar trailer film perlu dilakukan agar diperoleh trailer yang
disukai oleh penonton sehingga minat penonton untuk menonton film tersebut dapat
meningkat. Penelitian ini bertujuan mengembangkan model Natural Language
Processing (NLP) dengan metode word embedding FastText dan arsitektur Gated
Recurrent Unit (GRU) untuk analisis kecenderungan sentimen komentar di dalam
trailer film Oppenheimer. Berdasarkan hasil pelatihan dan evaluasi diperoleh model
TextBlob VADER sebagai model terbaik dengan accuracy bernilai 0,93. Hasil
analisis kecenderungan sentimen pada masing-masing trailer film menggunakan
model ini menghasilkan pengamatan bahwa trailer 2 lebih diminati penonton
dibandingkan trailer 1. Selain itu, penelitian ini juga telah berhasil mengembangkan
aplikasi web app sederhana untuk membantu proses analisis sentimen. In today's internet era, trailers are commonly uploaded on social media platforms like YouTube where viewers can easily access them and express their opinions. Sentiment analysis for a movie trailer is needed in order to create a trailer that is liked by the viewer so that it can increase the viewer’s interest in watching the movie. This research aims to develop a Natural Language Processing (NLP) model using the FastText word embedding method and Gated Recurrent Unit (GRU) architecture for analyzing sentiment tendencies in comments on the Oppenheimer’s movie trailers. Based on the training and evaluation results, the TextBlob VADER model was found to be the best model with an accuracy of 0,93. The sentiment trend analysis on each movie trailer using this model showed that trailer 2 was more favored by viewers compared to trailer 1. Additionally, this research successfully developed a simple web app to assist in the sentiment analysis process. |
URI: | http://repository.ipb.ac.id/handle/123456789/153582 |
Appears in Collections: | UT - Computer Science |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
cover_G6401201087_a34597afafac4fcb9f8aacf05e1d8973.pdf | Cover | 389.39 kB | Adobe PDF | View/Open |
fulltext_G6401201087_16a023c05900417fac50882460f5b6ab.pdf Restricted Access | Fulltext | 2.22 MB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.