Christian, Marcellio Aurel (2026) ANALISIS SIKAP DAN ISU TERHADAP PROGRAM MAKAN BERGIZI GRATIS PADA MEDIA SOSIAL X MENGGUNAKAN INDOBERT DAN BERTOPIC. Undergraduate thesis, UPN Veteran Jawa Timur.
|
Text (cover)
22082010019-cover.pdf Download (664kB) | Preview |
|
|
Text (BAB1)
22082010019-bab1.pdf Download (116kB) | Preview |
|
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (BAB2)
22082010019-bab2.pdf Restricted to Repository staff only until 25 May 2029. Download (336kB) |
|
|
Text (BAB3)
22082010019-bab3.pdf Restricted to Repository staff only until 25 May 2029. Download (262kB) |
|
|
Text (BAB4)
22082010019-bab4.pdf Restricted to Repository staff only until 29 May 2029. Download (982kB) |
|
|
Text (BAB5)
22082010019-bab5.pdf Download (10kB) | Preview |
|
|
Text (DAFTAR PUSTAKA)
22082010019-daftarpustaka.pdf Download (158kB) | Preview |
|
|
Text (LAMPIRAN)
22082010019-lampiran.pdf Restricted to Repository staff only until 25 May 2029. Download (27kB) |
Abstract
The Free Nutritious Meal Program is a government policy that has attracted considerable public attention on social media X. Discussions about this program contain various responses, including support, criticism, and descriptive information. This study aims to examine the stance of X users toward the Free Nutritious Meal Program by applying IndoBERT, identify emerging issues through BERTopic, and present the analytical results through a web-based visualization system. The dataset was collected from social media X from January 6, 2025, to October 6, 2025, using the keywords “Program Makan Bergizi Gratis” or “MBG”. The data were processed through several stages, including filtering, labeling, text preprocessing, classification modeling, topic modeling, and model implementation into a web application. The system also incorporates OCR as an additional feature to extract text from images when available. The evaluation results indicate that the best IndoBERT performance was achieved using normalized data with a 70:30 split and class weight, resulting in an accuracy of 0.8910 and an F1-Macro score of 0.8703. The classification results show that the Pro class was the most dominant, followed by Neutral and Contra. Issue modeling with BERTopic produced 12 actual topics excluding outliers, with a coherence score of 0.6367 and a stability score of 0.7774. These findings indicate that the combination of IndoBERT and BERTopic can describe stance tendencies and emerging issues related to the Free Nutritious Meal Program.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | Q Science > Q Science (General) Q Science > QA Mathematics Q Science > QA Mathematics > QA76.6 Computer Programming T Technology > T Technology (General) |
||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Information Systems | ||||||||||||
| Depositing User: | Marcellio Aurel Christian | ||||||||||||
| Date Deposited: | 25 May 2026 08:37 | ||||||||||||
| Last Modified: | 25 May 2026 08:37 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/52506 |
Actions (login required)
 |
View Item |