Subjective Answer Sheet Evaluation

The personal answer sheet marking is required to increase the efficiency and accuracy in the educational systems. In this paper, we present a machine-based model with a primary element: text removal and marking. The text extension component appoints OCR through OpinCV and Tesseract, in addition to preprocessing methods such as gracekel conversion and sound eradication, overcomes such issues such as handwriting styles and noise layouts. The marked component uses the bias classification trained on the corpora labeled, and includes obscure -matching methods for the evaluation of the meaningful equality of the students' answers of the model's answers. Designed using backnd, flask and pirate base, supports real-time handling of data with educational platforms and easy integration. This system proposes a solid, skewable and user-friendly approach to mark the individual's answer in educational institutions.Keywords: fuzzy matching, Flask, Gaussian Naive Bayes, machine learning, subjective response evaluation, automation, semantic similarity, and educational technology.