Summary: | 碩士 === 國立臺灣科技大學 === 資訊管理系 === 105 === The present research is dedicated to the problem of automatic emotion detection from text. A lot of knowledge is left to be discovered yet in this challenging and fast-evolving field and the research outcome may be valuable for different areas of the human activity. Successful emotion detection from text might contribute to creating a friendlier and more comfortable digital environment, as well as overcome communication difficulties, consequently boosting people’s emotional intelligence.
We tried to approach this problem from the interdisciplinary point of view, starting from defining the object of detection, emotions, taking a closer look at social networks, as a source of data, and moving on to test and compare several emotion detection strategies: lexical-based, machine learning based and hybrid ones.
|