Social networking sites have flooded the Internet with posts containing shared opinions, moods, and feelings. This gave rise to a new wave of research to develop algorithms for emotion analysis on social data. As the necessity to understand the posts of people grow past the boundaries of a country or region, the need to analyze social data in different languages grows. Traditional emotion classifiers require extracting high dimensional feature representations, which become computationally expensive to process and can misrepresent or deteriorate the accuracy of a classifier. In this project, we propose an unsupervised graph-based algorithm to extract patterns that bear emotion. By using the extracted patterns, a classification method can efficiently identify the emotions from posts. The full system works with different languages and domains. The experimented result demonstrate that the proposed approach bears desirable characteristics such as accuracy, generality, adaptability, and minimal supervision.
EmoTrend: Before and After [Golden Globe Award][Oscar]