What it does
The automated exam paper checking system will be developed using machine learning algorithms and natural language processing techniques. The system will use semantic analysis techniques to evaluate the meaning of the student responses rather than simply matching them to a pre-defined answer key. The system will identify the key concepts and ideas in the response and compare them to the expected answer, taking into account synonyms, paraphrasing, and other variations in language. The system will use advanced natural language processing techniques like sbert to create the sentence level embeddings which also have semantics associated on it and then finding cosine similarity between embeddings of expected answers and student answers to evaluate the meaning of student responses and compare them to the expected answer. To ensure the accuracy and reliability of the system, it will be tested extensively using a variety of exam questions and response types. The system will be compared to human markers and other automated exam checking systems to assess its performance and identify areas for improvement.
Log in or sign up for Devpost to join the conversation.