Research Article
Emotion Label Enhancement via Emotion Wheel and Lexicon
| Input: training sentence and its emotion label , weighting parameter λ, emotion lexicon L | | Output: emotion distribution of | | (1) | Extract all affective words from by looking up L | | (2) | for each | | (3) | Obtain all emotion labels of by looking up L | | (4) | Generate discrete Gaussian distribution for each according to (3) | | (5) | end for | | (6) | Generate discrete Gaussian distribution for according to (3) | | (7) | return |
|