Please use this identifier to cite or link to this item:
https://repository.uksw.edu//handle/123456789/30975
Title: | Analisis Sentimen Tweet Pengguna Twitter terkait Diabete menggunakan Metode Naive Bayes |
Authors: | Asyer, Afiyatar |
Keywords: | Analisis Sentiment;Tweet;Twitter;Diabetes;Naïve Bayes |
Issue Date: | 15-Oct-2023 |
Abstract: | Diabetes merupakan suatu penyakit yang berdampak serius terhadap kesehatan dan seluruh bagian tubuh penderitanya, banyak pandangan dari masyarakat terhadap penyakit diabetes yang menimbulkan perdebatan berkepanjangan. Twitter merupakan pilihan platform yang sering digunakan untuk memberikan opini publik, analisis sentimen di Twitter dapat memberikan gambaran persepsi pengguna terhadap diabetes baik secara positif, negatif ataupun netral untuk mengetahui literasi kesehatan terhadap persepsi obesitas pengguna platform Twitter. Analisis sentimen dilakukan dengan metode text maining dengan jumlah crawling sebanyak 26,038 dan menghasilkan sampel 4,635 data. Melalui algoritma Naïve Bayes hasil analisis sentimen didapatkan sentimen positif sebesar 22%, sentimen negatif sebesar 14%, dan sentimen netral sebesar 64%. Nilai akurasi dihasilkan sebesar 87%. Analisis sentimen tweet terkait diabetes menggunakan text maining lebih mengarah pada sentimen netral dibandingkan sentimen negatif dan sentimen positif. Nilai dari akurasi algoritma Naïve Bayes masuk dalam kategori “Good Classification”. Diabetes is a disease that has serious implications for the health and overall wellbeing of individuals, and there are diverse perspectives within society regarding this condition, leading to prolonged debates. Twitter is a popular platform where public opinions are often shared, and sentiment analysis on Twitter can provide insights into users' perceptions of diabetes, whether positive, negative, or neutral, to understand the health literacy regarding obesity among Twitter users. Sentiment analysis was conducted using text mining methodology, with a total of 26,038 crawled tweets, resulting in a sample of 4,635 data points. Through the Naïve Bayes algorithm, the sentiment analysis yielded 22% positive sentiment, 14% negative sentiment, and 64% neutral sentiment. The accuracy rate achieved was 87%. The sentiment analysis of tweets related to diabetes using text mining leaned more towards neutral sentiment than negative or positive sentiment. The accuracy of the Naïve Bayes algorithm falls within the category of "Good Classification". |
URI: | https://repository.uksw.edu//handle/123456789/30975 |
Appears in Collections: | T1 - Informatics Engineering |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
T1_672019061_Judul.pdf | 764.09 kB | Adobe PDF | View/Open | |
T1_672019061_Isi.pdf Until 9999-01-01 | 566.91 kB | Adobe PDF | View/Open | |
T1_672019061_Daftar Pustaka.pdf | 245.65 kB | Adobe PDF | View/Open | |
T1_672019061_Formulir Pernyataan Persetujuan Penyerahan Lisensi Noneklusif Tugas Akhir dan Pilihan Embargo.pdf Restricted Access | 377.28 kB | Adobe PDF | View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.