Building Conversational AI for Assamese Religious Text using Deep Learning
DOI:
https://doi.org/10.61707/n95fvk06Keywords:
Deep Learning, Conversational Agent, RNN, GRU, BiLSTMAbstract
This paper focuses on the use of Deep Learning techniques to build an AI Conversational Agent for Assamese religious texts. Assamese is spoken by the people of Assam which is a state of North East India. It is a language that has unique semantic and syntactic features. In order to handle the religious text in Assamese language specialized Natural Language Processing (NLP) task is required. In this paper we have presented the Deep Learning approach using bi-LSTM model to handle Assamese religious text and the specialized NLP functions that were developed to process the Assamese text. This study also involved the construction of dataset for Assamese language religious text which has been presented in this paper. We have also presented the performance and accuracy testing result of the proposed model and compared it with other Deep Learning model using matrices such as precision, recall, and F1-score. Our proposed model has achieved a significant improvement, reaching an impressive accuracy of 89.99%.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
CC Attribution-NonCommercial-NoDerivatives 4.0