SENTIMENT ANALYSIS OF KOREAN BLOG COMMENTS USING KOREAN CORPUS

For the final project for Machine Learning class, I assessed the capability of Korean corpus to assess the sentiment of the blog comments in Korean website. In 2011, Korea Advanced Institute of Science and Technology published Korean Emotional Word Dictionary, in which about 8000 words are evaluated based on the positivity and the negativity. I used this corpus to train a model that evaluated the sentiment of a phrase, and applied this model to the collection of 2000 blog comments from Korean Radical Anti-male Website, Womad (I do not endorse this organization in any way). The performance of the model was compared with the model that was trained with the actual comments.