CD Skripsi

ANALISIS SENTIMEN REVIEW PLAY STORE APLIKASI THREADS MENGGUNAKAN MODEL RECURRENT NEURAL NETWORK (RNN)

---

Threads, a social media application developed by Meta, has been a subject of
research interest since its release in October 2023. This study conducts sentiment
analysis using a Recurrent Neural Network (RNN) model to understand users'
opinions and sentiments regarding various aspects of their experience with the
Threads application. Sentiment analysis is a crucial technique in natural language
processing that enables the identification of positive, negative, or neutral
sentiments from text related to the application. The RNN model was chosen for its
ability to handle the complexity and relationships among words in lengthy text.
Previous research has demonstrated the effectiveness of RNNs in sentiment analysis
contexts with significant accuracy. Training results showed low accuracy on
training data (52.91%) and validation data (54.35%), indicating underfitting.
During testing on unseen data, the model achieved an accuracy of 58.38%, with
precision of 34.08%, recall of 58.38%, and an F1 score of 43.04%. Based on these
findings, the study suggests adopting advanced techniques such as employing more
complex network architectures or advanced natural language processing
techniques to enhance sentiment analysis accuracy on Threads user data.
Keywords: Threads, Natural Language Processing, Sentiment Analysis, RNN

---

Ketersediaan

Informasi Detail

Judul Seri

-

No. Panggil

2003111129

Penerbit

Pekanbaru : Universitas Riau – FMIPA – ILMU KOMPUTER., 2024

Deskripsi Fisik

-

Bahasa

Indonesia

ISBN/ISSN

-

Klasifikasi

2003111129

Tipe Isi

-

Tipe Media

-

Tipe Pembawa

-

Edisi

-

Subjek

ILMU KOMPUTER

Info Detail Spesifik

-

Pernyataan Tanggungjawab

guntur

Versi lain/terkait

Lampiran Berkas

Komentar

---

Anda harus masuk sebelum memberikan komentar