Fuzzy Sentiment Analysis for Hand Gesture Interpretation in Words Expression

抄録

This paper will discuss the interpretation of hand gestures using words with the fuzzy sentiment analysis approach in the tea-tasting experiment. In the experiment, participants were asked to drink tea, give a hand gesture, and choose three words that could interpret the hand gesture. The results show that hand gestures cannot be categorized as fixed sentiment labels, considering the meaning can differ depending on each preference. Generally, traditional sentiment categorizes labels as positive, negative, or neutral, which fails to capture human preferences. Therefore, we propose that the fuzzy sentiment model be used to solve the traditional problem, which provides a more flexible approach. Our findings show that interpreting hand gestures such as positive is not always positive but can be neutral-positive and vice versa. The results show that hand gestures alone are insufficient to determine sentiment; word expression is required to capture sentiment's different and complex aspects.