情緒對於敲擊鍵盤影響之研究

Abstract

自動化之情緒辨識技術是建立智慧型系統中非常重要的一環，它能使電腦更加人性化，做出較正確的回應。近年來，一個新穎的情緒辨識技術被提出，即利用在標準鍵盤上敲擊按鍵之特徵來辨識情緒。然而，對於此現象本身，先前的研究非常的有限。本研究以兩個實驗驗證了在不同情緒狀態下敲擊鍵盤特徵之差異以及對於情緒如何影響到敲擊鍵盤的假設。前人文獻提及情緒會影像到專注度，因此此研究假設情緒影響到敲擊鍵盤的原因是透過專注度。此研究之實驗利用臉部表情回饋假設作為情緒激發方法以及”嘟”聲作為專注度控制方法，並讓受測者鍵入指定的數字字串。 實驗結果顯示在25位受測者中，有23位在正向情緒與負向情緒時敲擊鍵盤特徵呈現明顯差異。再者，19位受測者滿足情緒影響專注度進而影響敲擊鍵盤之假設。除此之外，此論文亦將此技術做出應用，建立一套在數位學習環境中可即時對老師回報學生情緒狀態之智慧型監控系統(LSAM)，此系統之優點在於低成本(Low-cost)與可延展性(Scalable)。 此篇論文提供了將敲擊鍵盤特徵發展於情緒辨識系統之合理證據以及情緒影響敲擊鍵盤之可能生理路徑與原因。

The automatic emotion recognition technology is an important part of building intelligent systems to prevent the computers acting inappropriately and act more humanized. A novel approach for recognizing emotional state by their keystroke typing patterns on a standard keyboard was developed in recent years. However, there was very limited investigation about the phenomenon itself in the previous literatures. In this study, we conduct two controlled experiments to verify the difference on typing pattern between different emotional states and validate the hypothesis we made about how emotion affects typing. Since several studies have mentioned that emotion would influence attention, we assume that emotion affects keystroke typing pattern through attention. This study adapt facial feedback as the emotion eliciting method and a beep cue as the attention control method to conduct experiments which having the subjects typing the scheduled number sequence. The results prove that in overall 25 subjects, there are 23 subjects showing significant difference in typing features between positive and negative emotions. Further, 19 subjects show that emotion influenced keystroke typing through attention. Besides, this study also apply this technology in real world by proposing a low-cost scalable system on affective state monitoring (LSAM) which can report students’ emotion state to the tutor in real-time in e-learning environment. This paper provides an evidence for the reasonability about developing the technique of emotion recognition by keystroke and the possible reason about why we can recognize emotions by keystroke.