Recognizing Emotions in Human Computer Interaction: Studying Stress Using Skin Conductance

Liapis, Alexandros; Katsanos, Christos; Sotiropoulos, Dimitris P.; Xenos, Michalis; Καρούσος, Νίκος

doi:10.1007/978-3-319-22701-6_18

Cited by 45 publications

(21 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Research has linked GSR variation to stress and SNS arousal. As a person becomes more or less stressed, the GSR increases or decreases respectively (Hoogerheide, Renkl, Logan, Paas & van Gog, 2019;Liapis, Katsanos, Sotiropoulos, Xenos & Karousos, 2015, Smets et al, 2018. Additionally, research has also linked GSR readings to cognitive activity, claiming GSR responses increase when more cognitive load is experienced (Ikehara & Crosby, 2005;Nourbakhs et al, 2012;Setz et al, 2010;Shi et al, 2007, Yousoof & Sapiyan, 2013.…”

Section: Physiological Measures Of Cognitive Loadmentioning

confidence: 99%

Combining physiological data and subjective measurements to investigate cognitive load during complex learning

Larmuseau

Vanneste

Cornelis³

et al. 2019

FLR

View full text Add to dashboard Cite

Cognitive load theory is one of the most influential theoretical explanations of cognitive processing during learning. Despite its success, attempts to assess cognitive load during learning have proven difficult. Therefore, in the current study, students' self-reported cognitive load after the problem-solving process has been combined with measures of physiological data, namely, electrodermal activity (EDA) and skin temperature (ST) during the problem-solving process. Data was collected from 15 students during a high and low complex task about learning and teaching geometry. This study first investigated the differences between subjective and physiological data during the problem-solving process of a high and low complex task. Additionally, correlations between subjective and physiological data were examined. Finally, learning behavior that is retrieved from log-data, was related with EDA. Results reveal that the manipulation of task complexity was not reflected by physiological data. Nevertheless, when investigating individual differences, EDA seems to be related to mental effort.

show abstract

Section: Physiological Measures Of Cognitive Loadmentioning

confidence: 99%

Combining physiological data and subjective measurements to investigate cognitive load during complex learning

Larmuseau

Vanneste

Cornelis³

et al. 2019

FLR

View full text Add to dashboard Cite

show abstract

“…These results are in alignment with previous research in play technologies [10] that places the stress in the upper left area of the VA space. Hence, our exploration of the VA space started from defining R1 (6,3), a rather small region in the upper left corner of VA, which was iteratively expanded horizontally, vertically and diagonally as far as R9 (4,5). In this way, nine different stress regions were formed, as illustrated in Figure 2 and elaborated in Table 1.…”

Section: Resuls 31 Subjective Ratings Dataset and Va Regions Identifmentioning

confidence: 99%

“…The equation 2 is a root mean square error function and its value represents the signal's variability due to sampling rate frequency. The smoothing algorithm excluded nine signals due to signal degeneration (see [5] for details), thus our final dataset consists 142 records, which are used in subsequent analysis.…”

Section: Assignment Of Physiological Data To Regions Of Va Ratingsmentioning

confidence: 99%

See 1 more Smart Citation

Subjective Assessment of Stress in HCI

Liapis

Katsanos

Sotiropoulos

et al. 2015

Proceedings of the 11th Biannual Conference of the Italian SIGCHI Chapter

Self Cite

View full text Add to dashboard Cite

Thirty-one healthy participants performed five stressful HCI tasks (stimuli) while their skin conductance signals were monitored. The selected interaction tasks were most frequently listed as stressful by 15 interviewees, who were typical computer users. At the end of each task, participants expressed their perceived emotional experience using the dimensional Valence-Arousal (VA) rating space. The obtained VA ratings were used to define nine rectangular regions in the VA space, labeled as "stress". Next, five popular machine learning classifiers were employed to identify stress based on the associated skin conductance signals per tested region. Results showed sufficient cross-region stress recognition accuracy; L-SVM: Mean=62.3%, SD=9.3%. Our findings support that the VA scale may be used for stress self-assessments in the context of subtle interaction events, which are typically expected in most HCI tasks.

show abstract

“…Finally, if window width was above 100 points then signal was auto-excluded from the feature extraction process due to degeneration effect. Further details about the applied signal preprocessing method are also available in [13]. After the signal smoothing process, 21 statistical features (e.g., mean, median, min, max, standard deviation, minRatio and maxRatio) [6] were Table 1.…”

Section: Error = Sqrt(σ(xi -Xi-1)^2)/(2*n))mentioning

confidence: 99%

Stress recognition in human-computer interaction using physiological and self-reported data

Liapis

Katsanos

Sotiropoulos

et al. 2015

Proceedings of the 19th Panhellenic Conference on Informatics

Self Cite

View full text Add to dashboard Cite

This paper investigates gender differences in stress recognition in human computer interaction (HCI) for both objective (i.e., skin conductance) and subjective (i.e., valence-arousal VA ratings) metrics. To this end, 31 healthy participants, 18 females, performed five HCI tasks, while their skin conductance was recorded. These selected HCI tasks were the ones listed as the most stressful, by a group of typical computer users, who were involved in a face to face pre-experiment interview for the identification of stressful cases in computer interaction. After each task, participants rated their interaction experience using the valence-arousal scale. The collected data were split based on participants' gender. Skin conductance signals were analyzed using seven popular machine learning classifiers. In both groups the best stress recognition accuracy for all tasks was achieved by Linear Discriminant Analysis LDA; Males: Mean=94.8% and SD=1.5%, Females: Mean=98.9% and SD=0.3%. Self-reported data analysis revealed a significant difference on how both genders communicate their emotions using the arousal scale. Our findings tend to suggest that gender does not affect skin conductance data during subtle HCI tasks. However subjective ratings such as arousal of emotional experience must be utilized carefully.

show abstract

Recognizing Emotions in Human Computer Interaction: Studying Stress Using Skin Conductance

Cited by 45 publications

References 13 publications

Combining physiological data and subjective measurements to investigate cognitive load during complex learning

Combining physiological data and subjective measurements to investigate cognitive load during complex learning

Subjective Assessment of Stress in HCI

Stress recognition in human-computer interaction using physiological and self-reported data

Contact Info

Product

Resources

About