Search for collections on Undip Repository

ANALISIS SENTIMEN PADA LAYANAN STREAMING FILM DI TWITTER MENGGUNAKAN METODE LATENT DIRICHLET ALLOCATION DAN SUPPORT VECTOR MACHINE

ROYANI, Noorhanida and Widodo, Catur Edi and Warsito, Budi (2024) ANALISIS SENTIMEN PADA LAYANAN STREAMING FILM DI TWITTER MENGGUNAKAN METODE LATENT DIRICHLET ALLOCATION DAN SUPPORT VECTOR MACHINE. Masters thesis, UNIVERSITAS DIPONEGORO.

[img] Text
Cover dll.pdf

Download (1MB)
[img] Text
Bab 1.pdf

Download (25kB)
[img] Text
Bab 2.pdf

Download (630kB)
[img] Text
Bab 3.pdf
Restricted to Repository staff only

Download (434kB)
[img] Text
Bab 4.pdf
Restricted to Repository staff only

Download (959kB)
[img] Text
Bab 5.pdf

Download (18kB)
[img] Text
Daftar Pustaka.pdf

Download (139kB)
[img] Text
Lampiran .pdf
Restricted to Repository staff only

Download (336kB)

Abstract

Seiring dengan berkembangnya teknologi, memunculkan banyak platform online untuk streaming film. Streaming platform banyak digunakan masyarakat seperti netflix, disney+, hbo go, we tv, vidio. Selama ini permasalahan streaming banyak menimbulkan opini di masyarakat. Banyaknya perbandingan antar streaming platform menjadi perbincangan di media sosial yaitu twitter. Opini yang disampaikan pengguna tentang streaming platform berisi komentar positif dan komentar negatif yang mempengaruhi pengguna lainnya yang ingin menonton film. Oleh karena itu, opini tersebut dapat dijadikan analisis sentimen untuk mengungkapkan dan memberi informasi dari tanggapan masyarakat terhadap platform streaming. LDA dapat digunakan sebagai topic modelling dan SVM untuk klasifikasi. Pada tahapan pengambilan data dengan menggunakan tools framework scrapy dengan python, data diambil sebanyak 5.000 dilakukan preprocessing text dan count vectorize. LDA dapat mempresentasikan topik dan dokumen serta klasifikasi menggunakan SVM mendapatkan hasil bahwa komentar positif lebih banyak dari pada komentar negatif. Hasil evaluasi performa didapatkan nilai akurasi 91%, precision 90%, recall 91% dan f1 score 90%. Hasil penelitian ini mendapatkan bahwa jumlah komentar positif yang lebih dominan dari pada komentar negatif. Oleh karena itu, perlunya data yang seimbang dalam melakukan klasifikasi komentar streaming platform agar pada saat klasifikasi dapat memberikan hasil prediksi klasifikasi yang baik.
Kata Kunci: analisis sentimen, platform streaming, LDA, SVM

Along with the advancement of technology, many online platforms for streaming movies have emerged. Streaming platforms are widely used by the public, such as Netflix, Disney+, HBO Go, WeTV, and Vidio. Streaming issues have been a topic of discussion in society. The comparison between streaming platforms has become a conversation on social media, particularly on Twitter. User opinions on streaming platforms consist of both positive and negative comments, influencing other users who want to watch movies. Therefore, these opinions can be used for sentiment analysis to reveal and provide information about the public's response to streaming platforms. LDA can be used for topic modeling, and SVM for classification.In the data collection stage using the Scrapy framework with Python, 5,000 data points were gathered text preprocessing and count vectorize was performed. LDA can represent topics and documents, and classification using SVM yielded results that positive comments were more abundant than negative comments. The performance evaluation results showed an accuracy of 91%, precision of 90%, recall of 91%, and an F1 score of 90%. This research found that the number of positive comments is more dominant than negative comments. Therefore, there is a need for balanced data in classifying streaming platform comments so that the classification can provide good predictive results.
Keyword: sentiment analysis, platform streaming, LDA, SVM

Item Type: Thesis (Masters)
Uncontrolled Keywords: analisis sentimen, platform streaming, LDA, SVM
Subjects: Sciences and Mathemathic
Divisions: Postgraduate Program > Master Program in Information System
Depositing User: ekana listianawati
Date Deposited: 06 May 2024 08:05
Last Modified: 06 May 2024 08:05
URI: https://eprints2.undip.ac.id/id/eprint/22809

Actions (login required)

View Item View Item