經由臉部表情觀察網路成癮者的情緒反應

Abstract

網路如今已經深入每個人的生活，其存在帶給了人們全方位的便利性，但也因此帶來一些副作用。網路提供人們尋求娛樂和刺激的空間，卻造成人們過度的依賴，進而對人產生負面的影響，例如學業、事業、甚至身體和心靈上的健康等等，這樣的現象即被認為是網路成癮。自網路出現一直以來，許多研究就已經注意到網路成癮可能帶來的問題，並針對不同的層面探討，而情緒是其中之一。網路雖然可以成為人們抒發情緒的地方，卻也可能因為網路成癮的影響而產生一些負面的情緒反應。雖然一直以來都有針對網路成癮和情緒的相關研究，但多數是使用問卷形式，相較於問卷形式的研究，使用生理訊號更能表現出受測者的真實反應，而本研究使用了面部表情來觀察網路成癮者。為了分析面部表情，本研究使用了在研究領域上分析臉部表情最常被使用的方法Facial Action Coding System (FACS) 作為分析指標並且提出一個辨認FACS的演算方式，並將此方式在三個公開的資料庫上進行信度檢測。接著設計了一個實驗去觀察網路成癮者的情緒反應，希望藉此驗證網路成癮者在情緒的表現上是否有別於一般人。實驗中利用了影片作為刺激方式，以Chen Internet Addiction Scale (CIAS) 做為評估受測者網路成癮的程度，並以情緒問卷和面部表情觀察情緒方面的反應，而在本研究中所提出的辨認演算方式將用以分析受測者的面部表情。 實驗結果顯示，提出辨認FACS的方法在資料庫的測試中在信度和鑑別力上有著普遍良好的表現。網路成癮者在情緒刺激下會有較多的臉部表情產生，而在情緒問卷的評分上會傾向於隱藏自己的感覺，另外使用多線性分析後發現在生氣情緒刺激下所產生的臉部表情甚至可成為一個網路成癮的評估方式。整體而言，網路成癮者在情緒反應上會有別一般人，而這個結果使未來人們可以用情緒反應的方式來預測網路成癮。

The fast development of internet and information technology brings convenience to modern life, but also brings some side effects. Internet provides people a platform to seek happiness and excitement which may result in over-rely on internet, and leads people to suffer from negative consequences such as education, occupation, and even mental and physiological health. This phenomenon is regarded as "internet addiction". Many researchers had focus on internet addiction and its related issues, and emotion is one of the important topics. Although people use internet to express their feelings, the consequence of internet addiction still can lead to some negative emotion. Despite the fact that some studies had discussed the issue on internet addiction and emotion, the amount of research using physiological reaction is very limited. The physiological signal could reveal more reliable information about the subjects compare with the self-report. Therefore, facial expression as a physiological reaction and will be used in this research in order to investigate the internet addiction subjects. In order to analysis facial expression, we adapted Facial Action Coding System (FACS) as measurement, which is a most widely used method in the research of facial analysis. Next, this research proposed method to recognize FACS. Three databases were used to test the reliability of the proposed method. An experiment was conducted to investigate the emotion reaction on internet addiction subjects. Five films were used as emotion elicitation material and Chen internet addiction scale (CIAS) was used to evaluated the level of internet addiction. The emotion reaction will be investigated by emotional self-report and facial expression, and the facial expression will be analyzed by the proposed methods. The proposed recognition method shows acceptable reliability and discrimination on the testing database in general. The results in the experiment show that internet addiction subjects will show more facial expression than control group and the internet addiction subjects are more likely to inhibit their feelings on the score of emotional self-report. Further, the multiple linear regressions show that facial expression under the anger elicitation can become a potential predictor of internet addiction. To sum up, internet addiction subject shows different emotional reaction than others, which allow people to predict internet addiction by investigate their emotional reaction in the future.