NLP-Based Student Readability Assessment and Personalized Feedback System

Student writing assessment is important in academic learning, but manual evaluation of written responses can require considerable time and may provide limited information about specific writing difficulties. This paper presents an NLP-Based Student Readability Assessment and Personalized Feedback System for automated analysis of student English writing. The system allows users to enter text directly or upload PDF, DOCX, and TXT documents. Submitted content is analyzed for readability, grammar, spelling, vocabulary, sentence structure, and text statistics. Seven readability measures are used to estimate reading difficulty, while vocabulary richness, repeated words, difficult words, sentence length, and sentence type provide additional writing characteristics. The system combines these results to generate an overall writing-quality score and a feature-weighted proficiency estimate. Personalized suggestions explain detected weaknesses and recommend areas for improvement. Teacher-managed assignments with reference answers and student submissions are also supported. Analysis records are stored per user in a lightweight database.