Skip to content

Project Requirements

Academic, Technical, and Reproducibility Standards

This project satisfies the requirements of CS316 – Artificial Intelligence and Data Science while adhering to structured research, ethical AI principles, and reproducible engineering practices.

Academic Requirements

Course Alignment

The project fulfills the following CS316 criteria:

  • AI / ML Implementation

Implementation of embedding-based NLP models and SVM classifiers for severity prediction.

  • Evaluation Metrics

Performance assessed using accuracy, precision, recall, F1-score, and AUC.

  • Well-Documented Dataset

Synthetic longitudinal Arabic dataset with structured labeling.

  • Ethical Reflection

Privacy, fairness, transparency, and non-diagnostic boundaries addressed.

  • Reproducibility

Structured documentation of preprocessing, modeling, and evaluation.

  • IEEE-Style Paper

Research findings documented in formal academic format.

  • Source Repository

Clean, modular repository with installation instructions.

  • Technical Presentation

Slides and supporting materials summarizing methodology and results.

Documentation Expectations

The documentation includes:

  • Problem definition and research motivation
  • Literature awareness and related work
  • Dataset design and ethical considerations
  • Model justification and architectural choices
  • Experimental methodology
  • Results and error analysis
  • Limitations and future directions
  • Clear non-clinical disclaimers

Technical Requirements

Software Environment

  • Python 3.9 or later
  • pip package manager

Core Dependencies

  • numpy for numerical computation
  • pandas for dataset processing
  • scikit-learn for SVM classification and evaluation
  • matplotlib and seaborn for visualization
  • Embedding framework compatible with multilingual embeddings

Optional Deployment Dependencies

Required only for interactive prototype:

  • Streamlit for web interface
  • or Flask / FastAPI for API layer

Hardware Requirements

  • Standard CPU-based system
  • GPU not required
  • 8 GB RAM recommended

Reproducibility Standards

  • Fixed Random Seeds

Data splits and randomness controlled where applicable.

  • Dataset Integrity

Synthetic dataset stored separately from documentation.

  • Dependency Documentation

Requirements file ensures consistent environment recreation.

Ethical and Usage Constraints

  • Non-Diagnostic Use

The system must not be used for medical diagnosis or treatment decisions.

  • Human Oversight

Professional interpretation is required for any practical application.

  • Responsible Deployment

Real-world use requires governance, auditing, and compliance safeguards.

Evaluation Criteria

The project is assessed on:

  • Technical correctness and depth
  • Research quality and methodological rigor
  • Model performance and metric interpretation
  • Code organization and reproducibility
  • Communication clarity
  • Ethical awareness
  • Professional documentation standards

Refer to the official CS316 specification for detailed scoring breakdowns.