Currently, the discussion about hate speech in Indonesia is warm, primarily through social media. Hate speech is communication that disparages a person or group based on characteristics such as (race, ethnicity, gender, citizenship, religion and organization). Twitter is one of the social media that someone uses to express their feelings and opinions through tweets, including tweets that contain expressions of hatred because Twitter has a significant influence on the success or destruction of one's image.This study aims to detect hate speech or not hate Indonesian speech tweets by using the Bidirectional Long Short Term Memory method and the word2vec feature extraction method with Continuous bag-of-word (CBOW) architecture. For testing the BiLSTM purpose with the calculation of the value of accuracy, precision, recall, and F-measure.The use of word2vec and the Bidirectional Long Short Term Memory method with CBOW architecture, with epoch 10, learning rate 0.001 and the number of neurons 200 on the hidden layer, produce an accuracy rate of 94.66%, with each precision value of 99.08%, recall 93, 74% and F-measure 96.29%. In contrast, the Bidirectional Long Short Term Memory with three layers has an accuracy of 96.93%. The addition of one layer to BiLSTM increased by 2.27%.
The Indonesian government has enforced the New Normal rule in maintaining economic stabilization and also restraining the spread of the virus during the Covid 19 pandemic. This has become a hot topic of conversation on social media Twitter, many people think positive and negative.The research conducted is a representation of text mining and text processing using machine learning using the Naive Bayes Classifier classification method, the objective of the analysis is to determine whether public sentiment towards the New Normal policy is positive or negative, and also as a basis for measuring the performance of the TF-IDF feature extraction and N-gram in machine learning uses the Naive Bayes method.The results of this study resulted in the accuracy rate of the Naive Bayes method with the TF-IDF feature selection. The total accuracy was 81% with a Precision value of 78%, Recall 91%, and f1-Score 84%. The highest results were obtained from the use of the Naive Bayes and Trigram algorithm parameters, namely 84%, namely 84% Precision, 86% Recall, and 85% f1-Score. The Naive Bayes algorithm with the use of the trigram type N-Gram feature extraction shows a fairly good performance in the process of classifying public tweet data.
Kalimat sindiran atau sarkasme masih sering digunakan oleh kalangan publik untuk mengungkapkan maksud isi hati dan pikiran baik itu yang disampaikan secara langsng maupun tidak langsung. Sarkasme dilakukan untuk menyindir dan menyakiti hati seseorang dengan menggunakan bahasa atau kata yang didalamnya mengandung kata positif tetapi maknanya negatif sehingga sering sekali terjadi opini salah diklasifikasikan. Penelitian ini melakukan kombinasi antara proses sentimen analisis dengan deteksi sarkasme untuk pengklasifikasian opini yang terdapat pada Twitter. Proses analisis sentimen dilakukan dengan tahapan preprocessing dan ekstraksi fitur dan diklasifikan dengan menggunakan metode Support Vector Machine dilanjutkan dengan proses pendeteksian sarkasme yang dilakukan tahapan ekstraksi fitur dengan 4 set fitur yaitu sentiment related, punctuation-relate, lexical and syntactic, dan pattern-relate dan diklasifikasikan dengan menggunakan metode Random Forest Classifier. Hasil penelitian ini didapatkan peningkatan nilai rata-rata akurasi sebesar 16,61 %, nilai presisi sebesar 5,45 %, nilai recall sebesar 9,64% dan kenaikan nilai F1score sebesar 11,27% dengan jumlah data sebanyak 2.027 dengan rincian data dengan label positif berjumlah 1023, data dengan label negatif berjumlah 587 dan data dengan label netral berjumlah 462. Data sarkasme didapatkan dari tweet dengan label positif yang kemudian diberikan label sarkasme atau tidak sarkasme dan didapat hasil label dengan jumlah keseluruhan berlabel sarkasme berjumlah 354 dan tidak sarkasme berjumlah 669.
This research was conducted to apply the KNN (K-Nearest Neighbor) algorithm in conducting sentiment analysis of Twitter users on issues related to government policies regarding Online Learning. Research using Tweet data as much as 1825 Indonesian tweet data data were collected from February 1, 2020 to September 30, 2020. Using the python library, Tweepy. word weighting using TF-IDF, will be classified into two classes of sentiment values, positive and negative. After testing with K of 20, the highest accuracy results were obtained when K = 10 with an accuracy value of 84.65% with a precision of 87%, a recall of 86% f measure 87% and an error rate of 0.12% and a tendency was also obtained. public opinion on online learning tends to be positive.
Media sosial menjadikan masyarakat mengalami pergeseran perilaku baik budaya, etika dan norma yang ada, sehingga mereka dapat mengeluarkan opini-opini yang mereka miliki. Opini merupakan suatu pendapat dari pemikiran masayarakat mengenai suatu permasalahan yang sedang terjadi, saat ini Indonesia sedang dihadapkan oleh masalah mengenai virus Covid-19 yang memakan begitu banyak korban jiwa sehingga masyarakat mengeluarkan opini mereka mengenai virus tersebut dan kebijakan yang dilakukan pemerintah menghadapi virus tersebut.Penelitian ini bertujuan untuk mengetahui bagaimana sentiment publik terhadap kebijakan yang akan dilakukan pemerintah mengenai kebijakan lockdown ataupun pembatasan sosial berskala besar menggunakan metode Support Vector Machine denga ekstraksi fitur tf-idf dengan pengujian yang nantinya akan dilihat bagaimana nilai accuracy, precision, Recall dan F1-Score.Penggunaan metode Support Vector Machine dan ekstraksi fitur dengan tf-idf yang membagi kelas menjadi sentiment positif 68,75% dan negative 31,25% menghasilkan nilai accuracy sebesar 74%, precision sebesar 75%, recall sebesar 92% dan F1-Score sebesar 83%.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.