<![CDATA[In the digital era, with the widespread use of social media, the use of sarcasm in song lyrics often presents a unique challenge in the interpretation process. One issue is the difficulty in detecting sarcastic language due to its implicit nature and dependence on context. Conventional methods often fail to capture this complex language pattern, which may lead to misunderstandings. This research aims to develop a mobile-based application capable of identifying sarcasm in song lyrics using the Support Vector Machine (SVM) algorithm. The application is designed to detect sarcasm in song lyrics, which is often hard to identify accurately through traditional methods. The development process includes several stages, such as data collection, pre-processing song lyrics data, applying the Term Frequency-Inverse Document Frequency (TF-IDF) method, and feature extraction. A sarcasm keyword dataset containing 600 data points with sarcasm elements and a general song lyrics dataset without sarcasm elements were collected and used for machine learning model training. The processed data is then classified using Support Vector Machine (SVM), which categorizes the analysis results into two main categories: sarcasm and non-sarcasm. The proposed classification model demonstrates performance with an Accuracy of 98.14%, Sensitivity of 96.13%, Specificity of 100%, and MCC of 0.9645, indicating a strong ability to distinguish between sarcastic and non-sarcastic language. This research aims to enhance users' understanding of song lyrics, especially on social media, to reduce misunderstandings related to sarcasm. It is hoped that this research can contribute to the development of technology for understanding sarcastic language in song lyrics.]]>
