Prasetyantoro, Dikogigih (2021) PENERAPAN ALGORITMA K-MEANS CLUSTERING UNTUK MENGELOMPOKKAN DATA OPINI PENGGUNA TWITTER MENGENAI PANDEMI COVID-19. Undergraduate thesis, Universitas Stikubank.
![[thumbnail of HLM JDL]](/style/images/fileicons/application_pdf.png) |
PDF (HLM JDL)
Download (832kB) |
|
PDF (ABSTRAK)
Download (375kB) |
|
PDF (BAB I)
Download (501kB) |
|
PDF (BAB II)
Restricted to Repository staff only Download (591kB) |
|
PDF (BAB III)
Restricted to Repository staff only Download (732kB) |
|
PDF (BAB IV)
Restricted to Repository staff only Download (1MB) |
|
PDF (BAB V)
Restricted to Repository staff only Download (433kB) |
|
PDF (DAFTAR PUSTAKA)
Download (434kB) |
|
PDF (LAMPIRAN)
Restricted to Repository staff only Download (1MB) |
Abstract
Pandemi COVID-19 membuat masyarakat sering menggunakan media sosial untuk berkomunikasi jarak jauh saat melakukan karantina mandiri. Melalui media sosial daring twitter didapatkan suatu opini yang menjadi tren pembahasan. Banyaknya data opini membuat data menjadi tidak terstruktur dan tidak berkelompok. Hal ini menimbulkan permasalahan berupa sulitnya mengetahui opini mana saja yang dapat masuk ke kelompok tren pembahasan tertentu. Clustering dapat berguna dalam mengelompokkan data opini dan menjadi solusi masalah yang telah dipaparkan sebelumnya, dengan begitu kita dapat mengetahui tren pembahasan apa saja yang dibahas masyarakat saat pandemi COVID-19. Pada penelitian ini algoritma yang digunakan adalah algoritma K-Means dan davies bouldin index dalam menentukan jumlah cluster. Metode davies bouldin index dalam menentukan jumlah cluster terbaik menunjukkan 3 cluster merupakan jumlah yang optimal, dengan DBI score sebesar 7.6050425313519225. Hasil yang didapatkan dari 1350 data opini didalam 3 cluster ditemukan: 1070 opini di cluster 0, sebanyak 74 opini di cluster 1 dan 206 opini didalam cluster 2. ABSTRACT The COVID-19 pandemic has made people often use social media to communicate remotely when self-quarantining. Through the online social media, twitter, an opinion has become a trend of discussion. The large number of opinion data makes the data unstructured and ungrouped. This raises problems in the form of difficulty in knowing which opinions can fit into certain discussion trend groups. Clustering can be useful in classifying opinion data and being a solution to the problems that have been described previously, that way we can find out what trends are discussed by the community during the COVID-19 pandemic. In this study, the algorithm used is the K-Means algorithm and the Davies Bouldin Index in determining the number of clusters. The davies bouldin index method in determining the best number of clusters shows that 3 clusters is the optimal number, with a DBI score of 7.6050425313519225. The results obtained from 1350 opinion data in 3 clusters were found: 1070 opinions in cluster 0, as many as 74 opinions in cluster 1 and 206 opinions in cluster 2.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | NIM:17.01.55.0073 SKR.I.05.02.1990 |
| Uncontrolled Keywords: | Clustering, K-Means, Twitter, Pandemi COVID-19, Text Mining |
| Subjects: | Q Science > QA Mathematics > QA76 Computer software |
| Faculty / Institution: | Fakultas Teknologi Informasi > Program Studi Sistem Informasi |
| Depositing User: | Ani Mariawati |
| Date Deposited: | 11 Nov 2021 07:31 |
| Last Modified: | 11 Nov 2021 07:31 |
| URI: | https://eprints.unisbank.ac.id/id/eprint/8060 |
Actions (login required)
 |
View Item |
