<![CDATA[Detection of emotions from text is a problem of text classification based on the type of emotion. The availability of an emotional corpus plays an essential role in the detection of emotions. However, most corpus for emotional detection is available in English. This condition is a major problem when developing a system for detecting emotions from Indonesian texts. The emotional text corpus for Indonesian is very limited. Therefore, this research focuses on the development of Indonesian text emotional corpus. The development of such a corpus is the first step in the study of detecting emotions from the Indonesian text. The data source used to develop the corpus is a tweet. The annotation process is done automatically based on the hashtag (#) of emotions contained in a tweet with five types of emotions: happy, sad, angry, afraid, and love. This research produced an Indonesian emotional text corpus consisting of 500 complete tweets with emotional labels at the superordinate and basic levels. Emotion detection experiments were conducted to test the corpus using the Naive Bayes method. The accuracy of the experiments reached 82%, these results indicate that the corpus can be used in text emotion detection.]]>
Copyrights © 2019
