SWR302_SU26 AI-Powered Mental Health Support Platform
SWR302_SU26 AI-Powered Mental Health Support Platform
Mental health awareness has grown significantly in recent years, yet access to timely psychological support remains limited due to therapist shortages, high costs, and social stigma. Many individuals experiencing anxiety, stress, or early signs of depression are unable to receive help when they need it most.
To bridge this gap, the proposed solution is the development of an AI-Powered Mental Health Support Platform. This system uses conversational AI, emotion recognition, and clinical decision-support algorithms to provide users with accessible mental health screening, guided self-help programs, and real-time crisis detection. The platform intelligently escalates cases that require human intervention to licensed counselors while supporting the broader user population through AI-driven tools.
When a user registers, they complete an AI-driven onboarding assessment covering mood history, stress triggers, and lifestyle factors. Based on this data, the AI model generates a personalized Mental Health Profile and recommends a tailored self-help program (breathing exercises, CBT journaling prompts, mindfulness sessions). Users interact daily with an AI Companion Chatbot that tracks emotional trends, detects signs of deterioration, and adjusts program recommendations dynamically.
The platform features a Crisis Detection Engine that continuously monitors user inputs — text sentiment, response patterns, and self-reported mood scores — for high-risk indicators. When a risk threshold is exceeded, the system automatically transitions the user's case status to critical, sends an alert to an available counselor, and presents the user with emergency resources and a direct connection option. Counselors can review AI-generated session summaries, risk scores, and intervention history before engaging with the user.
A content management team curates the therapeutic content library, while the system administrator manages user accounts, configures AI model thresholds, monitors platform analytics (engagement rates, crisis response times), and ensures compliance with health data privacy regulations (e.g., HIPAA-equivalent standards). The platform integrates with external calendar services for counselor appointment scheduling and with emergency services APIs for high-severity escalations.
IMPORTANT NOTES
All answers must be written in the provided template file, in English, and must reflect this exam paper.
You must use Visual Paradigm or Drawio (all offline) to draw diagrams.
Use Windows + Shift + S to capture diagram images and Ctrl + V to paste them into the answer template.
Any answers containing unrelated content will receive ZERO points.
Failure to use the provided template file will result in ZERO points.
Question 1 (0.5 point)
Complete all parts of the first page of the Software Requirements Specification (SRS) template provided to you, based on the system described in this exam paper.
Question 2 (2.5 points)
Draw a Context Diagram to describe the scope of the AI-Powered Mental Health Support Platform described in this exam paper. Clearly identify all external actors and their interactions with the system.
Question 3 (4.0 points)
Based on your answer in Question 2, draw a Use Case Diagram for the system. Briefly describe all actors and use cases in the provided template.
Question 4 (3.0 points)
Based on your answers in
Question 2 and Question 3, draw a State Transition Diagram for the User Mental Health Case — one of the core AI-monitored entities in this system. The diagram must model all meaningful states a case passes through, from initial screening to resolution or escalation, including AI-triggered transitions. Your diagram must include at least five distinct states (e.g., New, Under Assessment, Active Program, At-Risk, Crisis, Escalated to Counselor, Resolved), clearly labeled transition events (including AI-detected triggers and user actions), and guard conditions where applicable.
Briefly describe each state and its business meaning in the provided template.
