The Automated Checking and Grading Tool is a Large Language model (LLM)-based system designed to assist university professors in evaluating Software Requirements Specification (SRS) and Software Design Description (SDD) documents. By leveraging a dataset of previously graded documents, the tool aims to provide consistent, objective assessments while significantly reducing manual effort. Using natural language processing (NLP) and Large Language Models (LLMs), the tool analyzes key aspects of documentation, such as clarity, completeness, adherence to business requirements, and proper use of design diagrams. Additionally, the system integrates a self-learning mechanism that continually improves its evaluation accuracy by incorporating feedback from professors. The final score combines the results of multiple evaluation criteria, delivering a detailed and fair assessment of each document’s quality. By automating this process, the tool improves grading consistency and reduces the workload for professors, allowing them to focus on higher-level tasks.

1. Automate Grading: Develop and implement an automated grading system for Software Requirements Specifications (SRS) and Software Design Descriptions (SDD) that can evaluate submissions based on predefined criteria, achieving at least 90% accuracy in grading within six months of deployment.

2. Provide Feedback: Create a feedback mechanism within the system that generates detailed, actionable feedback for each submission, allowing students to access their grades and feedback within 24 hours of submission, with professor approval.

3. Self-Learning Model: Integrate a self-learning algorithm that updates its grading criteria based on historical grading data and user feedback, resulting in a 20% improvement in grading consistency and accuracy over the first year of operation

1. Development of an Automated Grading Tool: Create a software application that automates the grading of Software Requirements Specifications (SRS) and Software Design Descriptions (SDD) using Large Language Models (LLMs) to evaluate documents based on predefined criteria.

2. Focus on Technical Documentation: Specifically target the grading of SRS and SDD documents, addressing the unique challenges posed by technical content, including both textual and diagrammatic elements.

3. Continuous Learning and Improvement: Implement a self-learning mechanism that enables the grading model to evolve over time by learning from new data and feedback, enhancing its accuracy and effectiveness.

4. User Training and Support: Provide training and support for faculty members to effectively utilize the tool, ensuring they can interpret the results and feedback generated by the system.

5. Feedback Mechanism: Develop a robust feedback system that not only grades the documents but also provides detailed insights into areas of strength and opportunities for improvement.

6. Evaluation Criteria: Define clear evaluation criteria for the grading process, focusing on aspects such as completeness, clarity, adherence to business needs, and overall structure of the documentation.

7. Future Scalability: Design the system with scalability in mind, allowing it to be adapted for other types of academic documentation in the future.