Analisis Sentimen Opini Masyarakat Indonesia di Media Sosial Twitter Pada Kebijakan Electronic Road Pricing Menggunakan Metode Support Vector Machine dan Random Forest

Dharmawan, Irfan Andre (2023) Analisis Sentimen Opini Masyarakat Indonesia di Media Sosial Twitter Pada Kebijakan Electronic Road Pricing Menggunakan Metode Support Vector Machine dan Random Forest. Other thesis, Institut Teknologi Sepuluh Nopember.

[thumbnail of 06211940000123-Undergraduate_Thesis.pdf] Text
06211940000123-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only until 1 September 2025.

Download (2MB) | Request a copy

Abstract

Jakarta, ibu kota Indonesia, merupakan salah satu dari 50 kota termacet di dunia menurut TomTom Traffic Index, yaitu berada di peringkat ke-29. Rata-rata warga Jakarta menghabiskan waktu selama 22 menit 40 detik untuk menempuh perjalanan sejauh 10 km. Karena kemacetan Jakarta semakin memburuk dibanding tahun-tahun sebelumnya, pemerintah DKI Jakarta berencana menerapkan Electronic Road Pricing (ERP) di beberapa ruas jalan pada awal tahun 2023 untuk mengurangi kemacetan dan mendorong penggunaan transportasi umum. Kebijakan tersebut telah menghasilkan dukungan dan penolakan dari publik. Media sosial Twitter merupakan wadah berdiskusi dan bertukar opini melalui tweet antar sesama penggunanya. Opini yang disampaikan dapat berupa opini positif maupun negatif yang kemudian dapat dilakukan analisis sentimen. Analisis pada penelitian ini adalah melakukan klasifikasi sentimen negatif dan positif terhadap kebijakan ERP menggunakan metode Support Vector Machine (SVM) dan Random Forest (RF) baik menggunakan oversampling dengan SMOTE maupun tanpa SMOTE. Perbandingan metode tersebut menggunakan k-fold cross validation dengan kriteria kebaikan klasifikasi accuracy menunjukkan bahwa hasil performa metode SVM Kernel RBF dengan SMOTE lebih baik daripada metode SVM Kernel Linier dan RF dalam mengklasifikasikan data teks sentimen masyarakat terhadap kebijakan ERP.
===================================================================================================================================
Jakarta, the capital of Indonesia, is one of the 50 most congested cities in the world according to the TomTom Traffic Index, ranking 29th. The average Jakartan spends 22 minutes and 40 seconds to travel 10 km. As Jakarta's congestion has worsened compared to previous years, the Jakarta government plans to implement Electronic Road Pricing (ERP) on some roads as early as 2023 to reduce congestion and encourage the use of public transportation. The policy has generated both support and opposition from the public. Social media Twitter is a place to discuss and exchange opinions through tweets between fellow users. Opinions conveyed can be positive or negative opinions which can then be analyzed sentiment. The analysis in this study is to classify negative and positive sentiments towards ERP policies using the Support Vector Machine (SVM) and Random Forest (RF) methods both using oversampling with SMOTE and without SMOTE. Comparison of these methods using k-fold cross validation with good classification accuracy criteria shows that the performance results of the RBF Kernel SVM method with SMOTE are better than the SVM Kernel Linear and RF methods in classifying public sentiment text data on ERP policies.

Item Type: Thesis (Other)
Uncontrolled Keywords: Electronic Road Pricing, Sentiment Analysis, Support Vector Machine, Random Forest; Electronic Road Pricing, Random Forest, Sentimen, SMOTE, Support Vector Machine, Twitter
Subjects: Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines.
Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
Q Science > QA Mathematics > QA76.F56 Data structures (Computer science)
Divisions: Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User: Irfan Andre Dharmawan
Date Deposited: 11 Aug 2023 08:50
Last Modified: 11 Aug 2023 08:50
URI: http://repository.its.ac.id/id/eprint/104578

Actions (login required)

View Item View Item