Fardliana, Arilza (2023) PEMBUATAN MODEL KLASIFIKASI DAN VISUALISASI PADA TWEET AKUN TWITTER SUARA SURABAYA. Undergraduate thesis, UPN Veteran Jawa Timur.
|
Text (Cover)
1635010010-cover.pdf Download (3MB) | Preview |
|
|
Text (Bab 1)
1635010010-Bab1.pdf Download (323kB) | Preview |
|
![[img]](http://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (Bab 2)
1635010010-Bab2.pdf Restricted to Registered users only until 24 July 2025. Download (585kB) | Request a copy |
|
|
Text (Bab 3)
1635010010-bab3.pdf Restricted to Registered users only until 23 July 2025. Download (376kB) | Request a copy |
|
|
Text (Bab 4)
1635010010-Bab4.pdf Restricted to Registered users only until 23 July 2025. Download (543kB) | Request a copy |
|
|
Text (Bab 5)
1635010010-Bab5.pdf Download (192kB) | Preview |
|
|
Text (Daftar Pustaka)
1635010010-DaftarPustaka.pdf Download (411kB) | Preview |
Abstract
Social media acts as a platform that is very effective as a medium for disseminating information, so it is widely used by online news platforms, one of which is Suara Surabaya, to provide up-to-date information, especially traffic jams, accidents, and criminal acts. However, with so many incidents occurring on the same day, the information provided is not clearly described, the information conveyed is unstructured and stacked. Currently, many methods have been used to create a classification model, one of which is the Naïve Bayes Classifier. In this study, we used this method to build a tweet classification model on the Suara Surabaya Twitter account. Tweet data is classified into three categories, namely traffic jams, accidents, and criminal acts. The stages of building this classification model are problem identification, literature study, data collection, location extraction using NER, pre-processing, model building, model evaluation, and data visualization. The best model creation results were obtained using Multinomial Naïve Bayes with an F1 score of 0.986 by performing a dataset comparison of 80:20 using the hold out method. Tweet data is visualized in the form of distribution maps and graphs where data visualization for each class of events is presented in the form of distribution maps according to known location points, and also in the form of diagrams to find out the number of events in each category.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | T Technology > T Technology (General) > T58.6-58.62 Management Information Systems | ||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Information Systems | ||||||||||||
| Depositing User: | Arilza Fardliana arilza | ||||||||||||
| Date Deposited: | 24 Jul 2023 04:32 | ||||||||||||
| Last Modified: | 24 Jul 2023 04:32 | ||||||||||||
| URI: | http://repository.upnjatim.ac.id/id/eprint/16060 |
Actions (login required)
 |
View Item |