DEVELOPMENT OF USER TRUSWORTHINESS MODEL IN TWITTER AND DEVELOPMENT OF INSTITUT TEKNOLOGI BANDUNG SOCIAL NETWORK ANALYSIS APPLICATION
There are many dangers in trusting misinformation. There's a need to filter trusted and untrusted information to avoid these dangers. One way to trust a piece of information is by checking the trustworthiness of the bearer of the information. Based on this background, a user trustworthiness...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/74161 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | There are many dangers in trusting misinformation. There's a need to filter trusted
and untrusted information to avoid these dangers. One way to trust a piece of
information is by checking the trustworthiness of the bearer of the information.
Based on this background, a user trustworthiness model was developed.
The SNA ITB application is a social network analysis developed by the SNA ITB
team. In this capstone final thesis, some features were added to enhance the
application capabilities. Furthermore, bugs and defects found in the application
were fixed.
User trusworthiness model was built by experimenting on some machine learning
architecture combination. Based on the experiment, the model with best
performance is the model which use CatBoost algorthim with standard scaling.
The model has f1-score of 0.9877.
The development of SNA ITB application used FastAPI, Celery, and PostgreSQL
in the backend. It also used Vue.JS for the frontend. The result of application
development is tested through functional testing, unit testing, and user acceptance
testing. |
|---|