Deep Learning Model for Sentiment Analysis in the Use of Informal Language and Slang On Social Media

The widespread use of social media has led to an explosion of information being conveyed in various forms of language, including informal language and slang. Sentiment analysis is important to understand the public opinion contained in this form of communication. This research explores the utilization of deep learning techniques to analyze sentiment in social media content, focusing on the challenges posed by informal language and slang. Deep learning models provide a more advanced and precise solution than traditional techniques for handling and analyzing texts that exhibit significant linguistic diversity. In this research, we designed and evaluated various deep learning architectures, such as Recurrent Neural Networks (RNNs), Long Short-Term Memory (LSTM), and Convolutional Neural Networks (CNNs), to classify sentiment as positive, negative, or neutral in social media data containing informal language and slang. The data collection used comes from popular social media platforms that contain comments, tweets, and other posts written in informal Indonesian. We also apply special text pre-processing techniques to handle slang and informal peculiarities, such as slang lexicons, abbreviations, and common spelling errors. The findings indicate that the LSTM model outperforms the other architectures, achieving the highest accuracy in sentiment classification for unstructured social media text. In conclusion, deep learning models have demonstrated their effectiveness in addressing the challenges of sentiment analysis in texts characterized by significant linguistic variation. This research makes an important contribution to the development of text analysis technology and opens opportunities for further research in this field.