AMELIA, NIKEN (2021) PUBLIC RESPONSE TO THE ECONOMIC IMPACT OF IMPLEMENTING GOVERNMENT POLICIES DURING THE COVID PANDEMIC USING MACHINE LEARNING APPROACHES. S1 thesis, Universitas Mercu Buana Jakarta.
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (HAL COVER)
01 Cover.pdf Download (377kB) |
|
Text (BAB I)
02 Bab 1.pdf Restricted to Registered users only Download (217kB) |
|
Text (BAB II)
03 Bab 2.pdf Restricted to Registered users only Download (284kB) |
|
Text (BAB III)
04 Bab 3.pdf Restricted to Registered users only Download (276kB) |
|
Text (BAB IV)
05 Bab 4.pdf Restricted to Registered users only Download (326kB) |
|
Text (BAB V)
06 Bab 5.pdf Restricted to Registered users only Download (149kB) |
|
Text (DAFTAR PUSTAKA)
07 Daftar Pustaka.pdf Restricted to Registered users only Download (137kB) |
|
Text (LAMPIRAN)
08 Lampiran.pdf Restricted to Registered users only Download (667kB) |
Abstract
The spread of COVID-19 has been rapidly increasing with a continuous rise in the number of victims globally. This has led to the implementation of various policies and rules by several countries, including Indonesia. The implementation of the PSBB (Pembatasan Sosial Berskala Besar) or Large-Scale Social Restriction and New Normal policies has caused several economic impacts with the circulation of several comments on social media. Therefore, this study aims to categorise people's opinions on social media, such as Twitter, into positive or negative classes using three algorithm classification methods, namely Bernoulli Naïve Bayes, Support Vector Machine, and K-Nearest Neighbor. It also compares the algorithm performance using the Hyperparameter Tuning technique to determine the best model with the highest accuracy. The results show that the use of the Hyperparameter Tuning technique improved the accumulation of accuracy value of the three models with a proportion of 80% training dataset and 20% testing dataset on dataset 1 and 2 of 18.50 and 10.50% with the best classification performance based on accurate values of SVM on dataset 1 and 2 of 96.50% and 97.25%, respectively. Visualisation modelling of word frequency in the form of word cloud shows that the implementation of the PSBB led to a decline in the economy. At the same time, the New Normal slowly recovered it although it harmed people’s health. Keywords: Sentiment Analysis, PSBB, New Normal, Economic Impact, Text Classification Penyebaran COVID-19 meningkat pesat dengan jumlah korban yang terus meningkat secara global. Hal ini mendorong diterapkannya berbagai kebijakan dan aturan oleh beberapa negara, termasuk Indonesia. Pelaksanaan kebijakan PSBB (Pembatasan Sosial Berskala Besar) dan New Normal memiliki beberapa dampak ekonomi dengan beredarnya berbagai komentar di media sosial. Oleh karena itu, penelitian ini bertujuan untuk mengkategorikan opini masyarakat di media sosial, seperti Twitter, ke dalam kelas positif atau negatif menggunakan tiga metode klasifikasi algoritma, yaitu Bernoulli Naïve Bayes, Support Vector Machine, dan K-Nearest Neighbor. Selain itu juga membandingkan performa algoritma menggunakan teknik Hyperparameter Tuning untuk menentukan model terbaik dengan akurasi tertinggi. Hasil penelitian menunjukkan bahwa penggunaan metode Hyperparameter Tuning meningkatkan akumulasi nilai akurasi ketiga model dengan proporsi dataset latih 80% dan dataset uji 20% pada dataset 1 dan 2 sebesar 18.50% dan 10.50% dengan kinerja klasifikasi terbaik berdasarkan nilai akurasi yaitu SVM pada dataset 1 dan 2 masing-masing adalah 96.50% dan 97.25%. Pemodelan visualisasi frekuensi kata dalam bentuk word cloud menunjukkan bahwa penerapan PSBB menyebabkan penurunan perekonomian. Pada saat yang sama, New Normal perlahan-lahan memulihkannya, meskipun membahayakan kesehatan orang. Kata kunci: Analisis Sentimen, PSBB, New Normal, Dampak Ekonomi, Klasifikasi Teks
Actions (login required)
 |
View Item |
