English
 
Help Privacy Policy Disclaimer
  Advanced SearchBrowse

Item

ITEM ACTIONSEXPORT

Released

Conference Paper

Emotive Stimuli-triggered Participant-based Clustering Using a Novel Split-and-Merge Algorithm

MPS-Authors
There are no MPG-Authors in the publication available
External Resource
Fulltext (public)
There are no public fulltexts stored in PuRe
Supplementary Material (public)
There is no public supplementary material available
Citation

Nath, S., Mukhopadhyay, D., & Miyapuram, K. (2019). Emotive Stimuli-triggered Participant-based Clustering Using a Novel Split-and-Merge Algorithm. In L. Dey, & S. Chaudhury (Eds.), CoDS-COMAD '19: Proceedings of the ACM India Joint International Conference on Data Science and Management of Data (pp. 277-280). New York, NY, USA: ACM Press.


Cite as: http://hdl.handle.net/21.11116/0000-0009-64CA-3
Abstract
EEG signal analysis is a powerful technique to decode the activities of the human brain. Emotion detection among individuals using EEG is often reported to classify people based on emotions. We questioned this observation and hypothesized that different people respond differently to emotional stimuli and have an intrinsic predisposition to respond. We designed experiments to study the responses of participants to various emotional stimuli in order to compare participant-wise categorization to emotion-wise categorization of the data. The experiments were conducted on a homogeneous set of 20 participants by administering 9 short, one to two minute movie clips depicting different emotional content. The EEG signal data was recorded using the 128 channel high-density geodesic net. The data was filtered, segmented, converted to frequency domain and alpha, beta and theta ranges were extracted. Clustering was performed using a novel recursive-split and merge unsupervised algorithm. The data was analyzed through confusion matrices, plots and normalization techniques. It was found that the variation in emotive responses of a participant was significantly lower than the variation across participants. This resulted in more efficient participant-based clustering as compared to emotive stimuli-based clustering. We concluded that the emotive response is perhaps a signature of an individual with a characteristic pattern of EEG signals. Our findings on further experimentation will prove valuable for the progress of research in cognitive sciences, security and other related areas.