Digital Handwritten Answer Sheet Evaluation System

Computerization of summary takes over the world. It requires a lot of time and energy from school teachers and university faculties. To counter this challenge, our project aims to optimize the assessment process by scanning handwritten responses by students in digital textbooks to meet teachers with predefined model responses. It's the purpose. This was achieved using state-ART technologies such as optical character detection (OCR), natural language processing (NLP), and machine learning algorithms. Advanced Bert bidirectional encoder displays from Transformer Models and Cosinus -similarity Algorithms are used for both accurate and effective assessment of student-provided responses. In contrast to response length, this project maximizes brand allocations in relation to the most important words to save time and effort from teachers and to promote a fair assessment process. Furthermore, this also encourages students to better understand the concept and provide accurate and accurate answers that contribute to the production of fair results and the same results. Keywords— Optical Character Recognition (OCR), Natural Language Processing (NLP), Machine Learning (ML), BERT (Bidirectional Encoder Representations from Transformers), Cosine Similarity, Text Classification, Tokenization, Artificial Nural Network (ANN), Sentence Embeddings.