Amanillah, Rahmatul (2025) Analisis Faktor Yang Mempengaruhi Kepuasan Peserta BPJS Kesehatan pada Rumah Sakit Wilayah Surabaya dengan Pendekatan Analisis Sentimen. Undergraduate thesis, UPN Veteran Jawa Timur.
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (COVER)
skripsi_scan.pdf Download (1MB) |
|
Text (BAB 1)
skripsi_scan (1).pdf Download (26kB) |
|
Text (BAB 2)
skripsi_scan (2).pdf Restricted to Repository staff only until 19 September 2027. Download (849kB) |
|
Text (BAB 3)
skripsi_scan (3).pdf Restricted to Repository staff only until 19 September 2027. Download (843kB) |
|
Text (BAB 4)
skripsi_scan (4).pdf Restricted to Repository staff only until 19 September 2027. Download (2MB) |
|
Text (BAB 5)
skripsi_scan (5).pdf Download (10kB) |
|
Text (DAFTAR PUSTAKA)
skripsi_scan (6).pdf Download (140kB) |
|
Text (LAMPIRAN)
skripsi_scan (7).pdf Restricted to Repository staff only until 19 September 2027. Download (110kB) |
Abstract
The National Health Insurance (JKN) program by BPJS has been present since 2014 to guarantee the basic health needs of the community. However, this program still faces various problems, such as limited quotas, less than optimal services, and constraints on facilities and policies. Thus, periodic studies are needed to assess the level of participant satisfaction with BPJS Kesehatan services. Therefore, researchers conducted sentiment analysis using the ServQual model in creating questionnaires and producing text data used in sentiment analysis. Sentiment analysis is carried out to absorb negative and positive sentiments in the data. The sentiment analysis process starts from data cleaning, data labeling, word weighting with TF-IDF, classification models with the support vector machine (SVM) algorithm, then model evaluation with configuration matrix, and finally visualization of positive and negative sentiment data with wordcloud. In addition, using the SMOTE method to handle data imbalance. The results of the analysis show that the SVM model with the RBF kernel is more consistent in classifying data from the 6 aspects studied, with the best model performance being with parameter C = 1 after SMOTE with an accuracy value of 90%, precision of 93%, recall of 88%, and F1_Score of 89%. Some of the findings that needed improvement included long queues, limited medical facilities, administrative inefficiencies, and inconsistencies in medical personnel services.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | T Technology > T Technology (General) > T385 Computer Graphics | ||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Data Science | ||||||||||||
| Depositing User: | Aman rahmatul amanillah | ||||||||||||
| Date Deposited: | 22 Sep 2025 02:25 | ||||||||||||
| Last Modified: | 22 Sep 2025 02:25 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/44198 |
Actions (login required)
 |
View Item |