top of page
  • Writer's pictureVinamra Rathi

AI For Education: Automating Essay Grading for Better Learning


In today's tech-driven world, education is evolving with the help of EdTech. One key area for innovation is essay grading, which is traditionally slow and inconsistent. To tackle this, we created an AI-powered essay evaluation tool to save time and help students improve their writing.


Understanding the Problem


Grading essays is one of the most time-consuming tasks for teachers, especially with the ongoing teacher shortage. This issue is worsened by student plagiarism and small errors, like missing parentheses or repeated words, that can be easily overlooked. These errors, if not caught, prevent students from learning from their mistakes. Therefore, a solution that can handle large volumes of essays efficiently is crucial.





How The AI Engine Helps Overcome the Problem


When teachers grade a student's essay, they look for features that match it to an A, B, or C grade based on their experience and specific assignment criteria. The AI engine is trained to imitate this grading process. Our AI-enabled grading system has broken down the criteria into three basic requirements to evaluate essays:


  • Thesis: Determines if the claim answers the prompt.

  • Claim: Checks if the claim restates key parts of the prompt.

  • Sentence Structure: Identifies run-on sentences.


A student can write their essay in response to a given prompt and receive feedback based on their input. With no restrictions, the student can continuously improve their essay using the AI grading engine's feedback. Once the student is satisfied with their essay, they can submit it, and the AI will grade it accordingly.




Designing the System


To achieve these goals, we needed a robust design. Our system architecture consists of the following key components:


  1. Large Language Models (LLMs): We deployed LLMs trained on a vast corpora of text data. These models are capable of understanding context, grammar, and writing style, which are crucial for evaluating essays.

  2. Criteria-Based Evaluation: We broke down the grading process into specific criteria, such as thesis clarity, argument support, and sentence structure. This allows for a more granular and objective assessment.

  3. Continuous Feedback Loop: The system provides continuous feedback, allowing students to revise their essays before final submission. This iterative process helps students learn and improve.


Results and Impact


The AI engine provides a continuous score, helping students quickly improve their writing. Students can review their responses and learn from the AI's feedback before submitting their final answers. This efficient grading method allows us to score many essays quickly, freeing teachers to focus on other tasks. Automation can significantly reduce teachers' stress, as they often spend up to 90 minutes a night grading papers.


However, there can be AI biases or instances where teachers feel the AI grading is inaccurate. In these cases, we allow users to review the content and manually adjust the scores.


Conclusion


Creating an AI-powered essay evaluation engine was challenging but rewarding. By combining large language models with a focus on fairness and accuracy, we developed a system that transforms essay grading. Our journey highlights AI's potential in education, offering a future where technology enhances learning for both students and teachers.


We are excited about the possibilities this technology holds and are committed to continuous improvement, ensuring our AI engine remains a reliable and valuable educational tool.



23 views0 comments

Комментарии


bottom of page