Abstract—Affective computing is a newly trend the main goal
is exploring the human emotion things. The human emotion is
leaded into a key position of behavior clue, and hence it should be
included within the sensible model when an intelligent system
aims to simulate or forecast human responses. This research
utilizes decision tree one of data mining model to classify the
emotion. This research integrates and manipulates the Thayer’s
emotion mode and color theory into the decision tree model, C4.5
for an innovative emotion detecting system. This paper uses 320
data in four emotion groups to train and build the decision tree
for verifying the accuracy in this system. The result reveals that
C4.5 decision tree model can be effective classified the emotion by
feedback color from human. For the further research, colors will
not the only human behavior clues, even more than all the factors
from human interaction.