Szczegóły publikacji
Opis bibliograficzny
Analysis of social media data using deep learning and NLP method for potential use as natural disaster management in Indonesia / Nadhila Nurdin, Krzysztof KLUZA, Maya Fitria, Khairun Saddami, Rika Sri Utami // W: COSITE 2023 [Dokument elektroniczny] : 2023 2nd international conference on Computer System, Information Technology, and Electrical Engineering : sustainable development for smart innovation system : 2–3 August 2023, Banda Aceh Indonesia : proceedings. — Wersja do Windows. — Dane tekstowe. — Piscataway : IEEE, cop. 2023. — Dod. ISBN: 979-8-3503-4307-6. — e-ISBN: 979-8-3503-4306-9. — S. 143–148. — Wymagania systemowe: Adobe Reader. — Bibliogr. s. 148, Abstr. — Publikacja dostępna online od: 2023-09-19
Autorzy (5)
- Nurdin Nadhila
- AGHKluza Krzysztof
- Fitria Maya
- Saddami Khairun
- Sri Utami Rika
Słowa kluczowe
Dane bibliometryczne
ID BaDAP | 148992 |
---|---|
Data dodania do BaDAP | 2023-11-09 |
Tekst źródłowy | URL |
DOI | 10.1109/COSITE60233.2023.10249849 |
Rok publikacji | 2023 |
Typ publikacji | materiały konferencyjne (aut.) |
Otwarty dostęp | |
Wydawca | Institute of Electrical and Electronics Engineers (IEEE) |
Abstract
During the last decade, the use of social media has exploded. One of the internet’s most popular and accessible social media sites is Twitter in which people can have micro-blog to post their views on any subject, called a tweet. Indonesia becomes one of country which has the most active user of social media. On the other side, Indonesia has experienced a great deal of natural disasters because it is located in a pacific ring of fire. It is needed effective disaster management used in Indonesia, especially with the help of social media data. Big data analysis obtained during turbulent and unorganized emergency situations is the perfect fit for effective management. During a disaster, it is vital to manage the proper decision to help people affected with their needs. This research investigated some methods used in social media analysis and then apply deep learning model for supervised tasks which classify the related and unrelated disasters on tweet datasets and Latent Dirichlet Allocation (LDA) for unsupervised learning which extracts the details category on the disaster Twitter data. The LSTM model architecture obtained a higher score accuracy than CNN. Moreover, the LDA topic modeling obtained potential results on the details topic from the datasets which could describe useful information to help disaster management.