Use this URL to cite or link to this record in EThOS:
Title: Enhancing electronic intelligent tutoring systems by responding to affective states
Author: Tao, Xiaomei
ISNI:       0000 0004 6346 8428
Awarding Body: Birmingham City University
Current Institution: Birmingham City University
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
The overall aim of this research is the exploration mechanisms which allow an understanding of the emotional state of students and the selection of an appropriate cognitive and affective feedback for students on the basis of students' emotional state and cognitive state in an affective learning environment. The learning environment in which this research is based is one in which students learn by watching an instructional video. The main contributions in the thesis include:  - A video study was carried out to gather data in order to construct the emotional models in this research. This video study adopted a methodology in qualitative research called "Quick and Dirty Ethnography"(Hughes et al., 1995). In the video study, the emotional states, including boredom, frustration, confusion, flow, happiness, interest, were identified as being the most important to a learner in learning. The results of the video study indicates that blink frequencies can reflect the learner's emotional states and it is necessary to intervene when students are in self-learning through watching an instructional video in order to ensure that attention levels do not decrease.  - A novel emotional analysis model for modeling student's cognitive and emotional state in an affective learning system was constructed. It is an appraisal model which is on the basis of an instructional theory called Gagne's theory (Gagne, 1965).  - A novel emotion feedback model for producing appropriate feedback tactics in affective learning system was developed by Ontology and Influence Diagram ii approach. On the basis of the tutor-remediation hypothesis and the self-remediation hypothesis (Hausmann et al., 2013), two feedback tactic selection algorithms were designed and implemented. The evaluation results show: the emotion analysis model can be used to classify negative emotion and hence deduce the learner's cognitive state; the degree of satisfaction with the feedback based on the tutor-remediation hypothesis is higher than the feedback based on self-remediation hypothesis; the results indicated a higher degree of satisfaction with the combined cognitive and emotional feedback than cognitive feedback on its own.
Supervisor: Jackson, Mike ; Ratcliffe, Martin ; Niu, Qinzhou Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available
Keywords: G400 Computer Science