Date
Publisher
arXiv
Automatic Essay Scoring (AES) assigns scores to student essays, reducing the grading workload for instructors. Developing a scoring system capable of handling essays across diverse prompts is challenging due to the flexibility and diverse nature of the writing task. Existing methods typically fall into two categories: supervised feature-based approaches and large language model (LLM)-based methods. Supervised feature-based approaches often achieve higher performance but require resource-intensive training. In contrast, LLM-based methods are computationally efficient during inference but tend to suffer from lower performance. This paper combines these approaches by incorporating linguistic features into LLM-based scoring. Experimental results show that this hybrid method outperforms baseline models for both in-domain and out-of-domain writing prompts.
Study at a glance
- Question: Can adding linguistic features to AI improve automatic essay scoring across different writing prompts?
- Setting and data: Not stated in the abstract
- Method: Compares an AI essay scoring model combining linguistic features against baseline models across different writing prompts
Why it matters
Automatic essay scoring reduces instructor grading workloads, but scoring essays across diverse prompts is challenging. Existing approaches face a tradeoff between resource-intensive training and large language models that suffer from lower performance.
Key findings
- Combining linguistic features with large language models scored student essays more accurately than baseline models across both familiar and new writing prompts.
- Large language models can score student essays with high computational efficiency, but they tend to suffer from lower scoring accuracy.
- Traditional feature-based scoring approaches often achieve higher accuracy on student essays, but they require resource-intensive training.
Summary drafted with AI from the paper’s abstract and may contain errors. See the full paper for details.
What is the application?
Who is the user?
Who age?
Why use AI?
Study design
