NEXT WORD PREDICTION AND PARAPHRASING USING NATURAL LANGUAGE PROCESSING

Abstract: The use of online tools for a faster generation of essays or sentences has always been very important. We propose a one-destination website that provides users with the facilities to be able to choose between Essay generation, next-word prediction, or paraphrasing. The tools used typically help users not just with typing but also with grammatical errors, better sentence construction, and quicker more efficient outputs. Our project uses GPT, web scraping, and transformers. Pre-trained transformers have attained a state-of-the-art performance for various NLP tasks. We mainly focus on providing users with a one-stop destination for their needs related to the topic. The website saves users time having to find results for their needs. Moreover, the site helps reduce plagiarism while still providing a good result for the user’s needs.

Keywords: Latent Semantic Analysis, Paraphrasing, Next word prediction, Natural language processing, BERT, transformers, LSTM, Character prediction.