Identifikasi Misinformasi dan Disinformasi tentang COVID-19 pada Twitter

Sari, Ayu Mutiara (2021) Identifikasi Misinformasi dan Disinformasi tentang COVID-19 pada Twitter. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

Text 05111740000149-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (2MB) | Request a copy

Abstract

Media sosial menjadi sebuah platform yang paling digemari pada masa pandemi COVID-19, karena masyarakat banyak menggunakan media sosial sebagai sarana memperoleh informasi dan bertukar informasi. Tetapi, informasi yang beredar di media sosial belum tentu benar. Tidak jarang informasi yang beredar adalah informasi yang salah dan bertujuan untuk membuat kekacauan di masyarakat. Tentunya jika terus dibiarkan, penyebaran informasi yang salah ini dapat menghambat penanganan COVID-19. Untuk itu, mengidentifikasi konten-konten di media sosial yang berisi informasi yang salah perlu dilakukan secepat mungkin untuk mencegah informasi tersebut menyebar luas dan menghambat penanganan COVID-19. Pada penelitian ini dilakukan percobaan menggunakan model machine learning classifier guna memperoleh model yang bekerja baik dalam mendeteksi informasi yang salah pada media sosial Twitter. Sebelum dilakukan training model machine learning classifier, dataset diproses melalui tahap pre-processing dan feature extraction. Proses tuning parameter menggunakan GridsearchCV dengan jumlah cross-validation 5 folds juga dilakukan guna memperoleh parameter yang cocok bagi model machine learning classifier. Model diuji menggunakan dua jenis data uji untuk masing-masing model. Dari keseluruhan model yang diuji, Dari keseluruhan model klasifikasi yang diuji, Hasil klasifikasi model Linear Support Vector Machine menggunakan TF-IDF memperoleh nilai F1-Score terbaik dari model lainnya pada kedua data uji, yaitu 92,21% pada data testing 1, dan 93,33% pada data testing 2.
=====================================================================================================
Social media has become the most popular platform during the COVID-19 pandemic, because many people use social media for obtaining information and exchanging information. However, the information circulating on social media is not necessarily true. The information that is circulated could be wrong information and aims to create chaos in the community. The spread of misinformation could hinder the handling of COVID-19. Identifying the misinformation content on social media needs to be done as quickly as possible to prevent the spread of this information from spreading and hampering the handling of COVID-19. In this study, an experiment was conducted using a machine learning classifier model in order to obtain a model that worked well in detecting misinformations on Twitter. Before training the machine learning classifier model, the dataset goes through the pre-processing and feature extraction. The parameter tuning process using GridsearchCV with a cross-validation of 5 folds was also carried out to obtain suitable parameters for the machine learning classifier model. We used two types of test data for each model. Of all the models tested, of the overall classification models tested, the results of the classification of the Linear Support Vector Machine model using TF-IDF obtained the best F1-Score value from the other models in the both of testing data. Linear Support Vector Machine model using TF-IDF obtained F1-Score 92,21% in data testing 1, and 93,33% in data testing 2.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	COVID-19, GridsearchCV, Machine Learning Classifier, Twitter, Word Embedding. COVID-19, GridsearchCV, Machine Learning Classifier, Twitter, Word Embedding.
Subjects:	T Technology > T Technology (General) T Technology > T Technology (General) > T57.5 Data Processing
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55201-(S1) Undergraduate Thesis
Depositing User:	Ayu Mutiara Sari
Date Deposited:	04 Aug 2021 05:50
Last Modified:	04 Aug 2021 05:50
URI:	http://repository.its.ac.id/id/eprint/84786

Actions (login required)

View Item