Deep Learning-Based Summarized Notes Generation from Document

Machine learning has emerged as a critical tool in enhancing students’ academic success, particularly in the context of summarizing comprehensive notes and bolstering their information retention capabilities. In the domain of natural language processing (NLP), text summarization is a challenging task. For students, the web contains rapid information on the same topic, which makes it difficult for them to go through all the information, especially during examination time. Text summarization plays an important role in scattered data. The proposed method in this paper uses BERT model for summarization of text. During the fine-tuning process for text summarization, it takes in a document and generates a summary by selecting the most important sentences or phrases. This is done by using a scoring system that assigns weight to each sentence or phrase based on its relevance to the overall document. This approach uses deep learning technique to recognize the handwritten class note, if not in digital document form followed by summarization of document in efficient manner. The task has been divided into three stages, first one is to recognize hand written text by applying OCR technique, second using BERT to summarize the text, and then evaluating summarized document by using intrinsic evaluation method.