Skripsi

ANALISIS SENTIMEN MEDIA SOSIAL FACEBOOK TERHADAP KEMACETAN LALU LINTAS DI KOTA PALEMBANG MENGGUNAKAN ALGORITMA NAÏVE BAYES

---

This study aims to analyze public sentiment toward traffic congestion in Palembang City based on Facebook social media comments using the Naïve Bayes algorithm. Data were collected through web scraping of Facebook comments related to traffic conditions in Palembang and labeled using a lexicon-based approach into three sentiment classes: positive, negative, and neutral. The dataset consists of 1,469 comments divided into training, validation, and testing sets. The preprocessing stages include data cleaning, case folding, tokenization, stopword removal, stemming, and normalization. Text feature representation was performed using the TF-IDF method prior to the classification process. The results show that the model achieved an accuracy of 98.44% on the training data, 85.97% on the validation data, and 85.91% on the testing data. The high training accuracy indicates that the model successfully learned sentiment patterns, while the decrease in validation and testing accuracy is influenced by language variations, the use of informal expressions and local dialects, as well as ambiguous comments. The characteristics of Facebook comments, which are generally short texts containing explicit sentiment keywords, support the performance of the Naïve Bayes classifier. However, class imbalance and limitations in capturing sentence context prevent the model from achieving optimal performance.

---

Availability

Detail Information

Series Title

-

Call Number

T1901072025

Publisher

Indralaya : Prodi Sistem Komputer, Fakultas Ilmu Komputer Universitas Sriwijaya., 2025

Collation

xiv, 119 hlm.; ilus.; tab.; 29 cm

Language

Indonesia

ISBN/ISSN

-

Classification

006.312 07

Content Type

Text

Media Type

-

Carrier Type

-

Edition

-

Subject(s)

Prodi Sistem Komputer
Teknik Pengolahan Data

Specific Detail Info

-

Statement of Responsibility

MI

Other version/related

File Attachment

Comments

---

You must be logged in to post a comment