This study aims to classify and identify the meaning of sentiment language to reduce misconceptions or misinterpretations of sentiment language in Twitter comment section regarding the 2024 general election in Indonesia and presidential candidates: Anies Baswedan-Muhaimin Iskandar, Prabowo Subianto-Gibran Rakabuming, Ganjar Pranowo-Mahfud M. D., as well as legislative canditates and the course of election. This research is descriptive qualitative and uses neutral language processing (NLP) methods, namely sentiment analysis lexicon-based in the form of manual labelling through identification of lexical indications and machine labelling by utilizing the sentiment analysis site Monkeylearn in unpaid mode or simple analysis on the surface of lexical indication data without any computational data processing and application of semantic theory to analyze changes in the meaning field and taste value in the form of sarcastic language styles. The results of this study are the classification of negative sentiment language in as much as twenty data, positive as much as six data, neutral as much as one data as well as the identification of three common sarcasm language styles in negative sentiment, namely irony, satire, and sarcasm as well as the tendency of expressing negative sentiments by Twitter users regarding candidates and the course of the election with piercing meanings but delivered with subtle language along with high negative sentiment scores but not in harsh language in the form of rhetorical cynicism dominance.