Oasis: Online Analytic System for Incivility Detection and Sentiment Classification

Incivility detection is an important task to identify offensive language on online social media platforms. Sentiment analysis is an essential task of natural language processing to identify the emotions of given sentences. In this demonstration, we propose an online processing system, called Oasis, to perform incivility detection and sentiment prediction for short texts (e.g., tweets). Oasis offers several useful features including the CNN-LSTM classification models, incivility detection and sentiment analysis for real-time tweets and user-interested queries, and personalized search of trending topics in Twitter.