Klasifikasi Sentimen Masyarakat Terhadap Perkembangan E-Sport di Indonesia Melalui Twitter Dengan Metode Artificial Neural Network Dan Support Vector Machine

Habibullah, M Irfandi Habibullah Iswandi (2019) Klasifikasi Sentimen Masyarakat Terhadap Perkembangan E-Sport di Indonesia Melalui Twitter Dengan Metode Artificial Neural Network Dan Support Vector Machine. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06211440000121-Undergraduate_Theses.pdf Restricted to Repository staff only Download (1MB) | Request a copy

Abstract

E-Sport merupakan salah satu cabang olahraga yang masih berkembang dimana negara maju E-Sport merupakan salah satu bidang yang diminati oleh banyak orang. Di Indonesia E-Sport sudah mulai banyak diminati oleh masyarakat. Dengan adanya perkembangan E-Sport di Indonesia penelitian ini ingin mengetahui pendapat masyarakat mengenai apakah pekembangan E-Sport di Indonesia disambut positif oleh masyarakatnya ataukah akan adanya penolakan dari masyarakat dengan adanya pekembangan E-Sport yang sedang terjadi di Indonesia. Untuk mengetahui pendapat masyarakat mengenai E-Sport di Indonesia dilakukan analisis sentimen menggunakan metode ANN menggunakan fungsi aktivasi Sigmoid Biner pada prediksi data awal memiliki nilai akurasi , F1-Score dan AUC masing-masing sebesar 72,8%, 66,0% dan 65,6%. Kemudian performa klasifikasi data SMOTE memiliki nilai sebesar 84,2%, 84,4%, dan 84,4%.dan SVM dengan menggunakan kernel linear pada prediksi data awal memiliki nilai 72,6%, 66,0% dan 65,7%. performa klasifikasi data SMOTE memiliki nilai sebesar 84,6%, 84,6% dan 84,9%.dan kernel non-linier pada prediksi data awal memiliki nilai sebesar 74,1%, 70,2%, dan 71,5%.Kemudian performa klasifikasi data SMOTE memiliki nilai sebesar 84,4%, 83,9% dan 84,1%. serta melakukan visualisasi kata yang sering muncul dalam tweet menggunakan WordCloud untuk mengetahui kata-kata apa saja yang sering muncul pada masing-masing kelas sentimen. ==================================================================================================================================
E-Sport is one sport that is still developing where developed countries E-Sport is one of the areas of interest by many people. In Indonesia E-Sport has begun to be in great demand by the public. With the development of E-Sport in Indonesia, this study wanted to find out the opinion of the public about whether the development of E-Sport in Indonesia was positively welcomed by the community or whether there was rejection from the community due to the development of E-Sport that was happening in Indonesia. To find out the public opinion about E-Sport in Indonesia, a sentiment analysis was conducted using the ANN method using the Binary Sigmoid activation function in the prediction of the initial data having an accuracy value, F1-Score and AUC of 72.8%, 66.0% and 65, respectively. 6%. Then the SMOTE data classification performance has a value of 84.2%, 84.4%, and 84.4%. And SVM using linear kernels in the prediction of the initial data has a value of 72.6%, 66.0% and 65.7%. . SMOTE data classification performance has a value of 84.6%, 84.6% and 84.9%. And the non-linear kernel in the initial data prediction has a value of 74.1%, 70.2%, and 71.5%. Then the data classification performance of SMOTE has a value of 84.4%, 83.9% and 84.1% . as well as visualizing words that often appear in tweets using WordCloud to find out what words often appear in each sentiment class.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	E-Sport, ANN, SVM, Analisis Sentimen, Leksikon, Twitter, WordCloud
Subjects:	T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Mathematics and Science > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	M Irfandi Habibullah Iswandi Putra
Date Deposited:	23 Jul 2026 02:09
Last Modified:	23 Jul 2026 02:09
URI:	http://repository.its.ac.id/id/eprint/64884

Actions (login required)

View Item