Book GPT: An Innovative PDF Querying Tool

Abstract

Access to information that is timely and reliable is becoming increasingly important in today's fast- paced environment. A common format for transmitting information, PDFs frequently contain crucial data that must be retrieved and analyzed. Yet, it can be time-consuming anderror-prone to manually search through PDFs for the necessary data. In this paper, we describe a hybrid strategy that makes use of the capabilities of Natural Language Processing (NLP) methods to read and produce precise responses from PDFs. We employ k-Nearest Neighbor to locate the data points that are most pertinent to a particular query and Universal Sentence Encoder to convert sentences into fixed-length numerical vectors. We also incorporate the cutting-edge language model Open AIGPT-3 top roduce text that is similar to that of a human being for increased accuracy. By creating the interactive user interface, our tool will enable the users to either upload a pdf or to provide an appropriate URL to ask a query. Our method makes it possible to quickly and accurately extract answers from PDFs in real-time applications, such as those seen in the information retrieval, legal ,and health care sectors[1].