Discourse “Ndasmu” on twetter after the 2024 Presidential Election Debate in Indonesia Topic Modeling and Sentiment Analysis

Abstract

The purpose of this study is to examine the conversation that transpired on Twitter following the Indonesia 2024 presidential election debate, with particular attention on the phrase "Ndasmu." To better comprehend the speech patterns and attitudes that surface in online interactions, this research use topic modeling and sentiment analysis. The following Twitter data collection steps are part of this research methodology: To retrieve relevant tweets, use the Tweepy library to access the Twitter API. Access can only be obtained using a Twitter API key. Pre-processing Text: Eliminating superfluous characters, links, and mentions from the text data. Tokenizing text means breaking it up into individual words or phrases. eliminating stop words. Topic Modeling: Non-negative Matrix Factorization (NMF) or Latent Dirichlet Allocation (LDA) are two methods used for topic modeling. Two well-liked Python packages for topic modeling are Scikit-learn and Gensim. Sentiment analysis can be done with TextBlob or NLTK. While TextBlob makes things easier, NLTK offers more customization options. Visualization: Producing word clouds for subjects and sentiment distribution, among other visual representations of the data, requires the use of tools like Matplotlib or Seaborn. It is anticipated that the study's findings would provide light on the prevailing themes in the "Ndasmu" discourse and the attitudes that accompanied the presidential debates. These findings can help us comprehend the dynamics of public opinion and how individuals react to important political occasions like debates for the presidential nomination.