The Impact of Linguistic Features on Emotion Detection in Social Media Texts

Emotions are an important aspect of human life, and scientific theories on emotions have been widely developed in various research fields such as philosophy, psychology, and neuroscience. In human-computer interaction, understanding emotions is also very important. Detecting emotions not only enables better decision-making, but is also useful in various contexts such as business, politics, and mental health. The focus on identifying emotions in text arises because emotions are often implied without explicit words. Through the analysis of grammar and sentence structure, text mining techniques enable the extraction of sentiments and emotions. Detecting and identifying emotions in text is important because it can be applied in a variety of fields, including decision-making, prediction of human emotions, product assessment, analysis of political support, and identification of depression. Text as textual data is an important source of information due to its ability to convey human emotions. In this research, emotion detection uses the Naïve Bayes method, with attribute weighting to improve accuracy using count vector. This classification approach allows grouping text into six emotion categories: happy, sad, fear, love, shock, and anger. The Naïve Bayes method was chosen for its reliability in classifying data based on conditional probabilities. Thus, this research provides a deeper understanding of understanding and managing emotions in the context of social media. The data classification results yield precision, recall, F1-Measure, and accuracy values.