The influence of fake accounts on sentiment analysis related to COVID-19 in Indonesia

Pratama, Rivanda Putra; Tjahyanto, Aris

doi:10.1016/j.procs.2021.12.128

Cited by 18 publications

(14 citation statements)

References 11 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The results showed the influence of fake accounts, which can reduce the performance of sentiment classification. Experimental results with both algorithms also prove that the Support Vector Machine algorithm has better performance than the Nave Bayes algorithm, with the highest accuracy value of 80.6% (Pratama & Tjahyanto, 2021).…”

Section: Literature Reviewmentioning

confidence: 78%

Sentiment Analysis Of Tourist Reviews Using K-Nearest Neighbors Algorithm And Support Vector Machine

Sari

Hermanto

Defriani

2023

SinkrOn

View full text Add to dashboard Cite

After Indonesia was awarded as a country with extraordinary natural charm, many foreign tourists came to Indonesia. According to the records of the Central Bureau of Statistics for 2020, approximately 5.47 million foreign tourists entered Indonesia. With the large number of foreign tourist visits, the need for tourist attractions is increasing, but finding information is now not difficult. One source of information for finding reviews of tourist attractions is TripAdvisor. On this website, there is a lot of information or reviews about various tourist attractions. However, the number of reviews makes tourists confused about identifying the quality of tourist attractions to be visited, so sentiment analysis needs to be done. Sentiment analysis itself is a technique to extract, identify, and understand sentiments or opinions contained in a text. In this research, two classification methods will be used in sentiment analysis techniques, namely K-Nearest Neighbors (K-NN) and Support Vector Machine (SVM). Besides that, the object of this research will be to focus on the most popular tourist attractions in Indonesia according to Trip Advisor, namely Waterbom Bali, Mandala Suci Wenara Wana, Teras Sawah Tegalalang, Pura Tanah Lot, and Pura Luhur Uluwatu. The purpose of the research is to find out the results of accurate sentiment analysis for the five tourist attractions and compare the two algorithms used. and after testing, it was found that the Support Vector Machine algorithm is superior to the K-Nearest Neighbors algorithm.

show abstract

Section: Literature Reviewmentioning

confidence: 78%

Sentiment Analysis Of Tourist Reviews Using K-Nearest Neighbors Algorithm And Support Vector Machine

Sari

Hermanto

Defriani

2023

SinkrOn

View full text Add to dashboard Cite

show abstract

“…The model tested uses several algorithms, including SVM, random forest, and logistic regression. The evaluation process is carried out by training the model according to the data distribution, starting from the training data distribution, with testing data including 80:20, 70:30, and 60:40 [22]. The test results of the three algorithms produced varying levels of precision, recall, and accuracy, as seen in Table 4.…”

Section: Resultsmentioning

confidence: 99%

Topic Sentiment Using Logistic Regression and Latent Dirichlet Allocation as a Customer Satisfaction Analysis Model

Cahyo,

Aesyi,

Santosa

2024

J.INFOTEL

View full text Add to dashboard Cite

Buying and selling goods now is more interesting through e-commerce or marketplaces because of the ease of carrying out online transactions. Each transaction usually generates a response from the customer. The transaction response on the Shopee platform is still in paragraph form and needs to be more specific. Therefore, this research aims to build a model analysis of customer satisfaction using the best algorithm between support vector machine (SVM), random forest, and logistic regression. This research method uses sentiment classification with logistic regression because the logistic regression algorithm has the best accuracy, with an accuracy of 90.5. Meanwhile, the SVM algorithm achieved an accuracy of 90.4, and random forest reached 90.2. The three algorithms were tested three times, splitting data train:test at 80:20, 70:30, and 60:40. The best results were obtained by splitting data at 60:40. The best model is used to predict data without labels. The prediction produces 12,844 positive sentiment comment data, 112 negative sentiment comment data, and 70 neutral sentiment comment data. The results of this research continued to topic modeling using latent dirichlet allocation (LDA) to generate a trending topic of customer satisfaction on sales products. Implications of discussing each trend topic can be used as a reference for improving products and services, especially in communicating with customers.

show abstract

“…Oleh karena VADER hanya bisa menganalisis teks dalam bahasa Inggris dan data yang dikumpulkan hanya data dalam bahasa Indonesia, maka dilakukan terjemahan data dengan bantuan library Google Trans. Pada analisis sentimen ini, polaritas dinilai melalui perhitungan berdasarkan bobot atau nilai masing-masing kata yang terdapat dalam kalimat [18]. Sebuah skor di atas 0 menandakan bahwa tweet tersebut mengandung sentimen positif, sementara skor di bawah 0 menunjukkan sentimen negatif, dan skor 0 mencerminkan sentimen yang bersifat netral.…”

Section: Analisis Sentimenunclassified

Preferensi Pengguna Twitter Terhadap Calon Presiden dan Wakil Presiden Indonesia 2024

Tresnani,

Yanto,

Hidayat

2024

jitsi

View full text Add to dashboard Cite

Pemilihan Presiden Indonesia 2024 semakin dekat dan marak diperbincangkan. Dinamika politik dan aktivitas yang dilakukan oleh pasangan calon presiden dan wakil presiden dapat memengaruhi preferensi masyarakat terhadap calon presiden dan wakil presiden. Seorang politisi pasti akan mempertimbangkan popularitasnya jika mereka mencalonkan diri sebagai presiden dan wakil presiden agar dipilih oleh masyarakat. Tujuan penelitian ini adalah melihat bagaimana preferensi atau kecenderungan masyarakat dalam memilih calon presiden dan wakil presiden Indonesia 2024, khususnya pengguna media sosial Twitter. Selain itu, penelitian ini untuk mengukur dampak informasi politik yang tersebar di Twitter dan menyediakan wawasan yang relevan serta signifikan bagi pengambil keputusan politik, tim kampanye pemilihan, dan analis politik untuk merancang strategi komunikasi yang efektif dengan memahami dinamika serta preferensi politik masyarakat di era digital. Penelitian ini menggunakan kombinasi metode text mining dan SAW (Simple Additive Weighting) untuk melihat kecenderungan atau pilihan masyarakat dalam memilih pemimpin negara berdasarkan data dari Twitter dengan beberapa hashtag dan kata kunci tertentu yang sering muncul di Twitter. Data yang terkumpul sebanyak 92.852 tweet yang kemudian dianalisis sentimennya. Data sentimen tersebut digunakan dalam proses pemeringkatan dengan metode SAW (Simple Additive Weighting). Hasil sentimen terhadap ketiga pasangan calon ini masih didominasi dengan sentimen positif. Berdasarkan hasil pemeringkatan menunjukkan bahwa pasangan calon nomor urut 2 yaitu Prabowo Subianto & Gibran Rakabuming Raka, mendapat peringkat pertama dengan nilai akhir 0.865, disusul pasangan calon nomor urut 1, Anies Baswedan & Muhaimin Iskandar, dengan nilai 0.846, dan terakhir pada peringkat 3 yaitu pasangan calon nomor urut 3, Ganjar Pranowo & Mahfud MD, dengan nilai 0.702.

show abstract

The influence of fake accounts on sentiment analysis related to COVID-19 in Indonesia

Cited by 18 publications

References 11 publications

Sentiment Analysis Of Tourist Reviews Using K-Nearest Neighbors Algorithm And Support Vector Machine

Sentiment Analysis Of Tourist Reviews Using K-Nearest Neighbors Algorithm And Support Vector Machine

Topic Sentiment Using Logistic Regression and Latent Dirichlet Allocation as a Customer Satisfaction Analysis Model

Preferensi Pengguna Twitter Terhadap Calon Presiden dan Wakil Presiden Indonesia 2024

Contact Info

Product

Resources

About