Emotion-Sensitive Virtual Assistants for Personalized Digital Learning Experiences

Hana Mori

Emotion-Sensitive Virtual Assistants for Personalized Digital Learning Experiences

Author(s)	Dr. Hana Mori
Country	United States
Abstract	Digital learning environments provide flexible access to educational resources, but many virtual assistants continue to personalize instruction primarily according to academic performance, previous responses, or content preferences while giving limited attention to the learner’s changing affective state. Confusion, frustration, boredom, anxiety, and engagement can influence persistence, attention, self-regulation, and willingness to continue challenging learning activities. This study proposes an Emotion-Sensitive Virtual Assistant for Personalized Digital Learning (ESVA-PDL) that combines learner-state estimation with adaptive pedagogical responses while maintaining human oversight, privacy protection, and uncertainty-aware intervention. The proposed architecture prioritizes non-biometric indicators, including voluntary learner self-report, dialogue sentiment, response latency, repeated help requests, error patterns, task abandonment signals, and interaction behavior. Based on the inferred learning state and confidence level, the assistant can adjust explanation depth, provide scaffolded hints, modify task difficulty, offer motivational prompts, recommend a short learning pause, or escalate persistent difficulties to a human educator. A design-science methodology was combined with evidence mapping and reproducible scenario-based simulation. The simulation contained 1,800 synthetic learning interactions distributed across five modeled affective states: engagement, confusion, frustration, boredom, and anxiety. Performance was compared between a conventional content-oriented virtual assistant and the proposed emotion-sensitive assistant using engagement, task completion, and learning-gain indicators. The proposed configuration achieved a mean simulated engagement index of 76.0 compared with 59.7 for the conventional assistant. Modeled task completion increased from 62.8% to 77.9%, while the mean learning-gain score increased from 4.4 to 6.5. The largest differences were observed during frustration, boredom, and anxiety, suggesting that adaptive emotional support may be particularly valuable when learners are at risk of disengagement. These results represent simulation-based validation rather than observations collected from actual students. The study concludes that emotion-sensitive learning systems should not attempt to imitate human emotion indiscriminately or continuously monitor students through intrusive sensing. Their educational value lies in using carefully interpreted learner-state information to select proportionate pedagogical support while preserving privacy, learner autonomy, teacher responsibility, and transparent system behavior.
Keywords	: emotion-sensitive learning; virtual assistants; personalized digital learning; affective computing; intelligent tutoring systems; learner engagement; adaptive learning; educational technology; human-centered learning; learner emotion
Field	Engineering
Published In	Volume 1, Issue 1, January-February 2020
Published On	2020-01-24

View / Download PDF File

About AJAMIR Fees & Payment Current Issue Publication Archive	Submit Research Paper Track Submission Status Publication Guidelines Publication Ethics Peer Review & Plagiarism	Join as a Reviewer Editors & Reviewers Reviewer Referral Program Get Reviewer Membership Certi.	Website/Journal Policies Usage Policy Content Policies Privacy Policy

Contact Us		+1-872-302-3851	editor@ajamir.com

American Journal of Advanced Multidisciplinary Innovation and Research

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Emotion-Sensitive Virtual Assistants for Personalized Digital Learning Experiences

Share this