This project explores the integration of AI chatbots into digital mental health interventions tailored for youths with Special Educational Needs (SEN), such as those with autism, learning disabilities, or emotional disturbances. The key innovation lies in addressing the unique challenges faced by SEN adolescents—stigma, communication barriers, and limited access to personalized mental health services—through AI-driven, non-judgmental companionship. Leveraging conversational AI technologies like natural language processing (NLP), the solution will offer scalable, 24/7 emotional support, skill-building exercises, and social interaction simulations.

The design integrates a mixed-methods approach: a 6-week randomized controlled trial (RCT) with 60 SEN participants (aged 12–18) will quantitatively assess mental health outcomes (e.g., anxiety via GAD-7, well-being via Warwick-Edinburgh Scale), while qualitative interviews and focus groups will capture user experiences. Backend analytics will track engagement patterns. Ethical safeguards include anonymization, simplified consent protocols, and bias-mitigated AI training data.

Potential impact includes empowering SEN youths to build autonomy, reducing reliance on overburdened traditional services, and informing inclusive AI design principles for vulnerable populations.

The inspiration stems from the exacerbated mental health crisis among SEN youths during COVID-19, where disrupted routines and limited access to tailored services worsened their isolation. Existing AI chatbots like Woebot show efficacy in general populations but lack customization for SEN needs. Research gaps in this domain (e.g., Johnson et al., 2020; Pittas & Papanastasiou, 2023) highlight the urgency for accessible, stigma-free interventions.

The hypothesis posits that SEN adolescents will engage more deeply with AI chatbots using informal, mood-adaptive dialogue, leading to measurable improvements in emotional well-being and self-efficacy. This belief is grounded in evidence that non-judgmental AI interactions can validate emotions and provide consistent support—critical for SEN individuals who struggle with human interactions. By partnering with SEN schools and advocacy groups, the project ensures relevance and cultural sensitivity, addressing systemic gaps in mental health education.

The solution employs NLP frameworks (e.g., GPT-based models) and user-centered design to create a SEN-specific chatbot. Core functionalities include emotion recognition, skill-building modules (e.g., social scenarios), and real-time mood tracking. Development requires collaboration with SEN educators for content validation, cloud infrastructure for scalability, and blockchain-like protocols for data confidentiality.

Market validation will involve pilot testing with a youth advisory board, iterative UI/UX refinements, and partnerships with SEN organizations to assess demand. Performance metrics include reduced anxiety scores, user engagement rates, and qualitative feedback on perceived companionship. To ensure positive UX, the interface will incorporate visual aids, simplified language, and adaptive response timing.

no

This project innovates by targeting an underserved population—SEN youths—through AI tools traditionally tested on general cohorts. Creative elements include hybrid quantitative-qualitative evaluation (e.g., Gioia’s method for thematic analysis) and inclusive design principles (e.g., bias audits of training data). Unlike existing chatbots, the solution adapts to sensory sensitivities (e.g., adjustable response speeds) and fosters agency by letting users guide conversations.

The integration of backend analytics with frontline caregiver insights ensures dynamic improvements, while ethical safeguards (e.g., pseudonymization, on-site counseling referrals) prioritize user safety. These innovations enhance effectiveness by aligning AI interactions with SEN-specific communication styles and emotional needs.

Scalability will be achieved through cloud-based deployment, modular design for easy updates, and collaborations with global SEN networks. Potential bottlenecks (e.g., diverse user profiles) will be mitigated via customizable chatbot personas and multilingual support.

Environmental sustainability is addressed through energy-efficient AI models and server optimization. Long-term engagement will rely on personalized content updates (e.g., new skill modules) and gamification elements (e.g., progress tracking). Continuous feedback loops with users and caregivers will ensure adaptability to evolving needs, while open-source frameworks will enable community-driven enhancements. Partnerships with policymakers will advocate for systemic adoption, ensuring lasting impact.

This project addresses systemic inequities in mental health support for SEN youths, a population disproportionately affected by stigma, communication barriers, and limited access to tailored services. By providing 24/7, stigma-free AI companionship, the solution enhances emotional well-being and fosters autonomy, directly advancing social goals of equity (SDG 10) and inclusive education (SDG 4). It empowers SEN adolescents to navigate social challenges independently, reducing their reliance on overburdened traditional systems and bridging gaps in care for marginalized groups.

Social impact will be measured through:
- Mental health metrics: Pre- and post-intervention reductions in anxiety (GAD-7) and improvements in well-being (Warwick-Edinburgh Scale).
- Engagement rates: Usage frequency, session duration, and skill-module completion.
- Qualitative outcomes: User-reported perceptions of companionship, self-efficacy, and reduced isolation from interviews and focus groups.
- Long-term indicators: Academic performance and social participation tracked via school partnerships.

To ensure responsiveness, the project employs a dynamic feedback loop:
- Community collaboration: Advisory boards comprising SEN youths, caregivers, educators, and advocacy groups will co-design updates and prioritize features.
- Data-driven adaptation: Backend analytics will identify usage trends (e.g., underutilized modules) to refine content and accessibility.
- Policy advocacy: Findings will inform guidelines for inclusive AI in public health, promoting systemic change.

By centering inclusivity in design (e.g., multilingual support, sensory-friendly interfaces) and maintaining ethical transparency, the solution aligns with global efforts to democratize mental health resources while safeguarding user dignity and agency.

Personal Information Collection Statement (PICS):
1. The personal data collected in this form will be used for activity-organizing, record keeping and reporting only. The collected personal data will be purged within 6 years after the event.
2. Please note that it is obligatory to provide the personal data required.
3. Your personal data collected will be kept by the LTTC and will not be transferred to outside parties.
4. You have the right to request access to and correction of information held by us about you. If you wish to access or correct your personal data, please contact our staff at lttc@eduhk.hk.
5. The University’s Privacy Policy Statement can be access at https://www.eduhk.hk/en/privacy-policy.

• I have read and agree to the competition rules and privacy policy.