Persuasive Computing Lab

Check out our latest publication with Oladapo Oyebode, PhD, Darren Steeves, Prof. Rita Orji, PhD

This paper presents, Recilify, an AI-driven and emotion-adaptive mobile health app that delivers personalized emotion regulation interventions based on users’ emotional states. Through a systematic design, development, and evaluation process, the app integrates persuasive strategies, emotion regulation techniques, and real-time emotion detection to improve mental well-being. Evaluation findings suggest that tailored, AI-powered interventions can enhance user motivation, engagement, usability, and support effective emotion regulation.

You can read the full paper here: https://lnkd.in/ghvwyucT

Dalhousie University, Dalhousie Computer Science

Check out our latest publication with Josteve Adekanbi, Japheth Mumo, Gladwin Irudayaraj, Olumide Thomas Adeleke, Ibukun Okunade, Prof. Rita Orji, PhD and Oladapo Oyebode, PhD

This paper presents HyperCare, an AI-driven, personalized, and adaptive persuasive technology for continuous hypertension prevention and management. HyperCare integrates multimodal sensing of health determinants, including sleep patterns, activity level, body weight, blood glucose, and blood alcohol content, with explainable artificial intelligence and persuasive strategies to deliver personalized, evidence-based interventions in real time. The system uses a retrieval-augmented generation based explainable large language model to analyze health data and provide adaptive recommendations. Through an expert evaluation involving clinicians and simulated hypertension-risk scenarios, HyperCare received high ratings for clinical relevance, transparency and explainability, persuasion, ethics, and usability, demonstrating its potential for continuous and personalized hypertension prevention and management.

You can read the full paper here: https://lnkd.in/ercxQcBJ

Dalhousie University

Check out our latest publication with Masud Imran, Gerry Chan, Oladapo Oyebode, PhD, Rebecca Moyer, Prof. Rita Orji, PhD

In this paper, we present KosmoWalker, a social space-themed, step-based mobile game that transforms walking into cooperative missions through goal setting and team challenges. Results showed that the prototype was perceived as engaging, persuasive, and easy to navigate, demonstrating the potential of social and collaborative exergame design. Overall, these findings illustrate the promise of social and collaborative exergame design and inform the development of a functional version for broader evaluation.

You can read the full paper here: https://lnkd.in/etNZqfim

Dalhousie University, Dalhousie Computer Science

Check out our latest publication with Olamiposi Olaiya, MSc, Gerry Chan & Prof. Rita Orji, PhD

This systematic review explores how Machine Learning (ML) is being combined with Augmented Reality (AR) and Virtual Reality (VR) in healthcare for disease diagnosis and treatment. We reviewed 43 studies published between 2014 and 2024 and found that the use of AR/VR-ML technologies has increased over time. Most studies focused on VR applications, commonly using Head-Mounted Displays and multiple ML algorithms together. Overall, AR/VR/ML systems show significant potential for supporting capabilities in diagnosing and treating many types of medical conditions; however, a large variability exists among the different studies due to differences in the design and evaluation of the studies.

You can read the full paper here: https://lnkd.in/eEUQq8Jj

Dalhousie University, Dalhousie Computer Science

Check out our latest publication with olusegun Latinwo and Prof. Rita Orji, PhD

This research presents RXGBoost, an explainable AI-based diagnostic framework that utilizes electroencephalography (EEG) recordings and demographic features to improve the accuracy of mental health disorder diagnoses. The framework integrates Boruta-XGBoost feature selection, adaptive class weighting, and SHAP-based interpretability to provide transparent and clinically meaningful diagnostic decisions. In evaluations across six mental health disorders, RXGBoost consistently outperformed deep learning and baseline machine learning models on multiple performance metrics, including accuracy, precision, sensitivity, specificity, and F1-score. The study also highlights the framework’s potential integration into IoT-based wireless body area network (WBAN) systems and wearable EEG platforms to support real-time, personalized, and explainable mental health care.

You can read the full paper here: https://lnkd.in/eQTbbrb6

Dalhousie University, Dalhousie Computer Science

Check out our latest publication with Mona Alhasani, Ali M. Alshareef, PhD, Prof. Rita Orji, PhD

This paper explores how gamified stress-management systems can become more effective through personalization. We examined data collected from 517 participants to investigate the relationships between gamification user types, coping strategies, and perceived stress. The results show that different user types prefer different coping mechanisms, while stress levels had little effect on these preferences. Based on these findings, we propose a dual-layered personalization framework that aligns coping strategies with gamification mechanics according to users’ HEXAD profiles. The study shows the importance of designing adaptive, context-aware interventions that support users’ psychological needs and intrinsic motivations.

You can read the full paper here: https://lnkd.in/eVWWJ8SR

Dalhousie University, Dalhousie Computer Science

Call for participation:
We are conducting a study on personalized, gamified stress-management apps.

Participation involves:
Short pre- and post-study surveys (~5–10 minutes each)
Using a mobile app (5–15 minutes/day) for 3 weeks or receiving stress-management resources
Optional short interview (10–15 minutes)

🎁 Receive a $15 CAD Amazon gift card upon completion.

🔗 Join here: https://lnkd.in/edeNUT_D

For inquiries, contact: Mona Alhasani ([email protected])

Dalhousie University, Dalhousie Computer Science

Check out our latest publication with Grace Ataguba (PhD), and Prof. Rita Orji, PhD

This chapter explores food and drug interactions, focusing on how foods and supplements react with drugs used to treat different health conditions. It reviews how machine learning (ML) algorithms are used to predict these interactions and support clinical decision-making. It also examines how much research has covered the pharmacokinetic impact of these interactions in empirical and ML-based studies. The main findings show that most interactions had a positive pharmacokinetic impact, while some were negative or inconclusive, and that deep neural networks achieved very high prediction accuracy.

You can read the full paper here: https://lnkd.in/dwdBHeKa

Dalhousie University, Dalhousie Computer Science

Check out our latest publication with Emeka Nwagu, Bilikis Banire, Grace Ataguba (PhD), Sandra Meier, Prof. Rita Orji, PhD

This review analyzes 46 peer-reviewed studies with 18,122 participants to uncover trends and identify gaps in problematic smartphone use (PSU) interventions. It evaluates interaction design strategies and their limitations in achieving sustained behavior change, considering the influence of intervention duration and user diversity. The review identifies key gaps and proposes eight research questions to advance the field. It offers a theoretical foundation and recommendations to improve the design, implementation, and evaluation of PSU interventions.

You can read the full paper here: https://lnkd.in/eqWr27sn

Dalhousie University, Dalhousie Computer Science

Check out our latest publication with Grace Ataguba (PhD) and Prof. Rita Orji, PhD

This study reviews how AI helps people develop healthy eating habits (i.e., through personalization based on each person's needs) and adapt its recommendations based on how each individual eats. Most of the AI research used to develop these tools appears to be effective; however, very few studies have followed participants long term to assess the long-term effectiveness of these tools. Some ethical issues regarding privacy and transparency were raised as well, and there is a need for improved research design in the future.

You can read the full paper here: https://lnkd.in/epctrEzQ

Dalhousie University, Dalhousie Computer Science