Pemodelan Penyebaran Informasi Demam Berdarah di Twitter (X) Menggunakan Social Network Analysis Berbasis Sentimen dan Spasial

Margaretha, Ribka Devina (2026) Pemodelan Penyebaran Informasi Demam Berdarah di Twitter (X) Menggunakan Social Network Analysis Berbasis Sentimen dan Spasial. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6026242014-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (8MB) | Request a copy

Abstract

Demam Berdarah Dengue (DBD) tetap menjadi tantangan kesehatan masyarakat yang signifikan di Indonesia, dengan lebih dari 257.000 kasus dan 1.400 kematian pada tahun 2024. Penelitian yang mengintegrasikan analisis sentimen, Social Network Analysis (SNA), dan analisis spasial-temporal untuk mengkaji penyebaran informasi DBD di media sosial masih sangat terbatas. Penelitian ini mengusulkan kerangka analisis terintegrasi yang diterapkan pada 5.715 tweet berbahasa Indonesia terkait DBD (Januari 2021-Desember 2025). Klasifikasi sentimen menggunakan IndoBERT mengungkapkan dominasi wacana netral (65,23%), diikuti negatif (20,31%) dan positif (14,45%). Empat jaringan interaksi berbasis sentimen dibentuk dan dianalisis, menunjukkan struktur yang sangat terfragmentasi (modularity di atas 0,98). Tenaga kesehatan profesional teridentifikasi sebagai aktor paling stabil berdasarkan betweenness centrality, sementara Kementerian Kesehatan RI memiliki in-degree tertinggi. Deteksi komunitas menggunakan algoritma Louvain mengidentifikasi 861 komunitas dengan pola diferensiasi yang jelas antarsentimen: komunitas negatif didominasi tenaga kesehatan dan platform tanya-jawab anonim, komunitas netral oleh institusi resmi, dan komunitas positif oleh kampanye pemerintah. Analisis spasial menunjukkan konsentrasi diskusi di koridor Jawa-Bali (Jakarta: 42,41%) serta variasi persepsi publik antarwilayah, sementara nalisis temporal mengungkapkan pola episodik dengan puncak berulang pada bulan Januari dan Juni, yang waktunya bervariasi antar tahun dan tidak sepenuhnya mengikuti pola musiman semata, serta korelasi positif yang signifikan meski lemah antara volume diskusi bulanan dan proporsi sentimen negatif (r = 0,31; p = 0,017). Temuan ini menunjukkan nilai integrasi dimensi sentimen, jaringan, spasial, dan temporal bagi analisis komunikasi kesehatan masyarakat, serta dapat menjadi acuan bagi instansi kesehatan dalam merancang strategi komunikasi yang lebih tepat sasaran.
=================================================================================================================================
Dengue fever remains a significant public health challenge in Indonesia, with more than 257,000 cases and 1,400 deaths recorded in 2024. Studies integrating sentiment analysis, Social Network Analysis (SNA), and spatial-temporal analysis to examine dengue-related information diffusion on social media remain particularly limited. This study proposes an integrated analytical framework applied to 5,715 Indonesian-language dengue-related tweets (January 2021-December 2025). Sentiment classification using IndoBERT revealed a predominantly neutral discourse (65.23%), followed by negative (20.31%) and positive (14.45%) sentiment. Four sentiment-based interaction networks were constructed and analysed, revealing highly fragmented structures with modularity scores exceeding 0.98. Healthcare professionals emerged as the most robustly influential actors by betweenness centrality, while the Ministry of Health held the highest in-degree values. Community detection using the Louvain algorithm identified 861 communities with clear sentiment-based differentiation: negative communities were dominated by healthcare professionals and anonymous Q&A platforms, neutral communities by official institutions, and positive communities by government-led awareness campaigns. Spatial analysis indicated geographic concentration of discussions along the Java-Bali corridor (Jakarta: 42.41%) alongside meaningful regional variation in public perception, while temporal analysis revealed an episodic pattern with recurring peaks in January and June, though peak timing varied across years and was not fully explained by seasonality alone. These findings demonstrate the value of integrating sentiment, network, spatial, and temporal dimensions for public health communication analysis, and may serve as a reference for health authorities in designing more targeted communication strategies.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	Demam Berdarah Dengue, Sentiment-based Social Network Analysis, Analisis Spasial, Analisis Temporal, Media Sosial, Dengue Fever, Sentiment-Based Social Network Analysis, Spatial Analysis, Temporal Analysis, Social Media
Subjects:	H Social Sciences > HM Sociology > HM742 Online social networks.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 59101-(S2) Master Thesis
Depositing User:	Ribka Devina Margaretha
Date Deposited:	29 Jul 2026 02:02
Last Modified:	29 Jul 2026 02:02
URI:	http://repository.its.ac.id/id/eprint/139059

Actions (login required)

View Item