Recognizing reader's affect using EEG data
Date of Publication
2017
Document Type
Master's Thesis
Degree Name
Master of Science in Computer Science
College
College of Computer Studies
Department/Unit
Computer Science
Thesis Adviser
Ethel Chua Joy Ong
Defense Panel Chair
Judith J. Azcarraga
Defense Panel Member
Nelson M. Marcos
Ethel Chua Joy Ong
Rafael A. Cabredo
Merlin Teodosia C. Suarez
Abstract/Summary
Emotion or act is known to play vital roles in rational and intelligent behavior, such as cognition and decision-making. Detecting or recognizing act can be done by analyzing physiological data or a combination of various physiological data. The current work presents a study on brainwaves or EEG signals, which are examples of physiological data, and their association to emotions while a person is reading literary action, an unexplored domain. EEG data from 32 participants were collected while they were reading a short story. These EEG signals were collected with the use of an Emotive Insight EEG headset, attached to the head of each participant while reading the story segments presented via the developed data collector tool. After which, features were extracted and dierent datasets were built according to sex, reading preference, and reading frequency proles. Decision Trees were used to establish baseline performance results, and these were able to classify the Hourglass of Emotion model and Emotions of Literary Response models. Support Vector Machines and Multilayer Perceptrons were trained on the same datasets to see if there is an increase in performance. Results show that they indeed yielded better performance results than DT, however, only by a small degree. Principal Component Analysis was used as an approach for feature selection, and results show comparable performance as opposed to using the base feature set of all EEG features with an averaged 5 margin of error.
Abstract Format
html
Language
English
Format
Electronic
Accession Number
CDTG007050
Shelf Location
Archives, The Learning Commons, 12F Henry Sy Sr. Hall
Physical Description
1 computer disc ; 4 3/4 in.
Keywords
Electroencephalography; Signal processing--Digital techniques; Machine learning
Recommended Citation
Kalaw, K. F. (2017). Recognizing reader's affect using EEG data. Retrieved from https://animorepository.dlsu.edu.ph/etd_masteral/5830