Rizky, Joy Lawa
Universitas Nusa Mandiri

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

Analisis Sentimen Media Sosial Youtube Kereta Cepat (Whoosh) Menggunakan Algoritma Bidirectional-LSTM Rizky, Joy Lawa; Gata, Windu
Progresif: Jurnal Ilmiah Komputer Vol 20, No 2: Agustus 2024
Publisher : STMIK Banjarbaru

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35889/progresif.v20i2.1958

Abstract

This study analyzes social media sentiment on YouTube regarding the high-speed train (Whoosh) using the Bidirectional-LSTM algorithm. The issue raised is the need for a deeper understanding of public perception of the high-speed train project, which can affect its acceptance and sustainability. The purpose of this paper is to evaluate the performance of the Bidirectional-LSTM algorithm in sentiment analysis compared to other algorithms. The method used involves collecting YouTube comment data, text preprocessing, and applying the Bidirectional-LSTM algorithm for sentiment classification. The parameters analyzed include accuracy, precision, and resilience to data variations. The research results show that the Bidirectional-LSTM algorithm achieves an accuracy of (0.86), which is significantly higher compared to the Multinomial Naïve Bayes algorithm (0.80), USE-Transfer learning (Tensorflow) (0.80), and Text Vectorization and Embedding (Tensorflow) (0.80). The conclusion of this study is that Bidirectional-LSTM is more effective and reliable in analyzing social media sentiment towards the high-speed train (Whoosh).Keywords: Sentiment Analysis; YouTube Fast Train (Whoosh); Bidirectional-LSTM. AbstrakPenelitian ini menganalisis sentimen media sosial YouTube terhadap kereta cepat (Whoosh) menggunakan algoritma Bidirectional-LSTM. Masalah yang diangkat adalah perlunya pemahaman yang lebih mendalam tentang persepsi publik terhadap proyek kereta cepat, yang dapat mempengaruhi penerimaan dan keberlanjutannya. Tujuan penulisan ini adalah untuk mengevaluasi performa algoritma Bidirectional-LSTM dalam menganalisis sentimen dibandingkan dengan algoritma lain. Metode yang digunakan melibatkan pengumpulan data komentar YouTube, preprocessing teks, dan penerapan algoritma Bidirectional-LSTM untuk klasifikasi sentimen. Parameter-parameter yang dianalisis meliputi akurasi, presisi, dan ketahanan terhadap variasi data. Hasil penelitian menunjukkan bahwa algoritma Bidirectional-LSTM mencapai akurasi (0.86) yang secara signifikan lebih tinggi dibandingkan dengan algoritma Multinomial Naïve Bayes (0.80), USE-Transfer learning (0.80), dan Text Vactorita-tion and Embedding (Tensorflow) (0.80). Simpulan penelitian ini adalah bahwa Bidirectional-LSTM lebih efektif dan andal dalam menganalisis sentimen media sosial YouTube kereta cepat (Whoosh).Kata kunci: Analisis Sentimen; Youtube Kereta Cepat (Whoosh); Bidirectional-LSTM;