Analisis Sentimen HateSpeech pada Pengguna Layanan Twitter dengan Metode Naïve Bayes Classifier (NBC)

Abstract

Twitter users in January 2023 increased by 27.4% over the prior-year period worldwide. One of the uses of Twitter as a social networking service is used to create HateSpeech. HateSpeech fosters hatred and hostility of certain individuals or groups (ethnic, religious, racial and inter-group) in the form of insults, defamation, blasphemy, provocation, incitement, and spreading false news. The solution to avoid divisions between factions that threaten the unity of the Indonesian nation needs to be done with sentiment analysis to group tweets that are included in HateSpeech or Non-HateSpeech. This study aims to analyze the HateSpeech sentiment of twitter service users with the Naïve Bayes Classifier method. The dataset used in this study was 5000 data processed with Python tools for the preprocessing stages (case folding, tokenization, stopword removal, normalization, dan stemming), Labeling, Term Weighting weighting with Term Frequency (TF), and Inverse Document Frequency (IDF). The K-Fold Cross Validation technique is carried out to validate data by dividing training and testing data into 3 test scenarios, namely 70% training data and 30% testing, 30% training data and 70% testing, and 50% training data and 50% testing of the Naïve Bayes Classifier Method. Based on evaluation using Confusion Matrix, the highest accuracy of 80%, precision 100%, recall 80%, and F1-Score 89% was obtained in the training data testing scenario of 70% and testing 30%.

Linear Kernel Optimization of Support Vector Machine Algorithm on Online Marketplace Sentiment Analysis

“…The explanation of the stages of the text preprocessing process starts from the case folding stage or changing the uppercase letters in a sentence so that the sentence form is uniform. The next flow is tokenization, which is changing the sentence into a smaller word form [21]. The next flow is stopword removal, which aims to eliminate common words in the text that often appear but do not provide important or relevant information in text analysis.…”

Analisis Sentimen Pengguna Instagram Pada Calon Presiden 2024 Menggunakan Algoritma Support Vector Machine

“…Analisis sentimen ini bertujuan untuk menganalisis opini, sentimen, dan emosi yang terdapat dalam dokumen atau data. Tugas pokok dari analisis sentimen yaitu untuk mengklasifikasikan sifat teks yang terdapat dalam kalimat atau pendapat (Murni et al, 2023).…”