Hidayat, Aulia Rahma and Putra Pandu Adikara, S.Kom., M.Kom. and Sigit Adinugroho, S.Kom., M.Sc (2020) Klasifikasi Hoaks Kesehatan di Media Sosial menggunakan Support Vector Machine. Diploma thesis, Universitas Brawijaya.
Abstract
Pada zaman sekarang ini media sosial menjadi salah satu alat komunikasi setiap orang dalam berbagai umur. Sebagai salah satu alat komunikasi yang dipakai oleh setiap orang tak jarang dijumpai berita-berita tidak jelas sumbernya atau berita Hoaks. Berita Hoaks mengenai kesehatan banyak tersebar di Media Sosial dan hal tersebut dapat memengaruhi kesadaran masyarakat akan pentingnya kesehatan. Pemisahan berita kesehatan yang benar dan tidak benar perlu dilakukan untuk menghindari hal tersebut. Proses pemisahan dilakukan dengan mengklasifikasikan berita kesehatan pada Media Sosial dengan metode Support Vector Machine dengan fitur Bag Of Words dan Lexicon Based Features. Data yang digunakan pada penelitian ini 80 berita yang didapatkan dari berbagai Media Sosial. Data kemudian dimasukkan dalam proses pre-processing untuk mendapatkan kata yang menunjukkan sebuah dokumen, kemudian dilanjutkan kedalam proses pembobotan kata menggunakan perhitungan TF-IDF. Hasil proses pembobotan kata dimasukkan pada proses inti yaitu perhitungan metode Support Vector Machine. Hasil pengujian parameter optimal didapatkan nilai gamma (γ) = 0,001, nilai lambda (λ) = 1, nilai epsilon = 0,000001, nilai degree (d) = 2 dan nilai maksimum iterasi = 30. Hasil evaluasi sistem menggunakan fitur Bag of Words dan Lexicon Based Features mendapatkan hasil yang baik dibanding dengan menggunakan salah satu fitur saja, hasil yang didapatkan dengan penggabungan fitur yaitu Accuracy sebesar 1; Precision sebesar 1; Recall sebesar 1; F-measure sebesar 1. Pengujian menggunakan K-fold Cross Validation juga dilakukan dengan nilai fold 10 dan didapatkan nilai rata-rata hasil Accuracy sebesar 0,6; Precision sebesar 0,68; Recall sebesar 0,47; F-measure sebesar 0,48.
English Abstract
In this day and age social media is one of the communication tools for people of all ages. As one of the communication tools used by everyone it is not uncommon to find unclear sources or Hoaks. Hoaks about health are widely spread on Social Media and this can affect public awareness of the importance of health. Separating true and untrue health news needs to be done to avoid this. The separation process is done by classifying health news on Social Media with the Support Vector Machine method with Bag of Words and Lexicon Based Features. The data used in this study 80 news obtained from various Social Media. The data is then entered in the pre-processing process to get the word that shows a document, then proceed to the word weighting process using TF-IDF calculation. The results of the word weighting process are included in the core process, namely the calculation of the Support Vector Machine method. Optimal parameter test results obtained gamma value (γ) = 0.001, lambda value (λ) = 1, epsilon value = 0,000001, degree value (d) = 2 and maximum iteration value = 30. The results of the system evaluation using the Bag of Words feature and Lexicon Based Features get good results compared to using only one feature, the results obtained by combining features namely Accuracy of 1; Precision of 1; Recall of 1; F-measure of 1. Testing using K-fold Cross Validation was also carried out with a fold value of 10 and obtained an average value of Accuracy results of 0,6; Precision of 0,68; Recall of 0,47; F-measure of 0,48.
Other obstract
-
| Item Type: | Thesis (Diploma) |
|---|---|
| Identification Number: | 0520150066 |
| Uncontrolled Keywords: | Klasifikasi, Hoaks, Support Vector Machine, Bag of Words, Lexicon Based Features, K-fold Cross Validation, Classification, Hoaks, Support Vector Machine, Bag of Words, Lexicon Based Features, K-fold Cross Validati |
| Subjects: | 000 Computer science, information and general works > 006 Special computer methods > 006.3 Artificial intelligence > 006.31 Machine learning |
| Divisions: | Fakultas Ilmu Komputer > Teknik Informatika |
| Depositing User: | Budi Wahyono Wahyono |
| Date Deposited: | 28 Feb 2021 15:11 |
| Last Modified: | 12 Apr 2023 06:53 |
| URI: | http://repository.ub.ac.id/id/eprint/183640 |
![[thumbnail of DALAM MASA EMBARGO]](http://repository.ub.ac.id/style/images/fileicons/text.png) |
Text (DALAM MASA EMBARGO)
0520150066-Aulia Rahma Hidayat.pdf Restricted to Registered users only until 31 December 2023. Download (2MB) |
Actions (login required)
 |
View Item |