Question Answering System using NLP and BERT
Shreya Acharya, K. Sornalakshmi, Bidisha Paul, Anshul Singh · 2022 3rd International Conference on Smart Electronics and Communication (ICOSEC) · 2022
This project proposes a simple and smart question answering system using Named Entity Recognition (NER) and Bidirectional Encoder Representations from Transformers (BERT) for communication purposes. Communication is a major factor that eases our day-to-day work. In this smart Question answering system, the system will create a customized dataset for the user based on the answers given by the user. Using NER, only the selective answers are taken, i.e., the most important answers, and then updated in the dataset. Based on that dataset, if a person then asks the system a question, the system will use BERT and CDQA (closed domain question answering) to fetch the best answer for the question asked. The resulting system will be an interactive one.