Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM66259
Volume:
10 |
7 |
Jul 2026
Pages:
1-3
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.
Department of Artificial Intelligence and Data Science
Anjuman-I-Islam Abdul Razzak Kalsekar Polytechnic Campus
Department of Artificial Intelligence and Data Science
Anjuman-I-Islam Abdul Razzak Kalsekar Polytechnic Campus
Manual grading of student essays is time intensive and often inconsistent, while most automated scoring systems merely return a numeric score without actionable feedback. This paper presents a dual module NLP framework that combines a Longformer based multi trait essay scoring model with a fine tuned Mistral 7B generative feedback engine. The scoring module predicts separate scores for content, organisation, language, and mechanics on a 0–10 scale. These rubric scores are then fed, along with the essay text, into a prompt conditioned large language model that generates a concise, personalised feedback paragraph highlighting strengths, weaknesses, and a concrete revision suggestion. The system is trained on the ASAP++ corpus of 12,978 student essays and evaluated with quadratic weighted kappa (QWK) for scoring, and teacher rated Likert scores for feedback quality. The Longformer scoring model achieves an average QWK of 0.78, outperforming a BERT base baseline (0.71). Human evaluators rated the generated feedback as “actionable” in 84 % of cases. The integrated pipeline reduces essay assessment turnaround from days to seconds, making it a practical tool for formative writing instruction.
KEYWORDS
This paper presented an integrated NLP system that performs multi‑trait essay scoring and generates personalised, rubric‑aware feedback using Longformer and Mistral‑7B. The system achieves strong agreement with human raters and produces feedback that teachers consider actionable. Future work will extend the framework to other writing genres, incorporate multi‑modal inputs (handwritten essays), and perform a longitudinal classroom study to measure impact on student writing outcomes.
1. Taghipour, K., Ng, H.T.: A Neural Approach to Automated Essay Scoring. In: Proc. EMNLP, pp. 1882–1891 (2016) 2. Beltagy, I., Peters, M.E., Cohan, A.: Longformer: The Long‑Document Transformer. arXiv:2004.05150 (2020) 3. Nguyen, H., Dang, K., Tran, T.: Automated Feedback Generation for Student Writing using Pre‑trained Language Models. In: Proc. ACL Workshop on Innovative Use of NLP for Building Educational Applications, pp. 183–192 (2022) 4. Jiang, A.Q., Sablayrolles, A., Mensch, A., et al.: Mistral 7B. arXiv:2310.06825 (2023) 5. Devlin, J., Chang, M.‑W., Lee, K., Toutanova, K.: BERT: Pre‑training of Deep Bidirectional Transformers for Language Understanding. In: Proc. NAACL‑HLT, pp. 4171–4186 (2019)
Publication Date
7/20/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
3
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...