Penerapan Neural Network pada Chatbot Tanya Jawab Ilmu Tajwid

Abstract

The use of chatbots is widely used in various fields such as economics, education, health, and so on. This research aims to develop and evaluate a chatbot as a learning tool for tajweed for Qur'an learners. The science of tajweed is very important because it learns how to read the Qur'an correctly according to the guidance of the Prophet Muhammad PBUH. The research method used is to develop a Neural Network for string matching to provide a response that matches the required answer. The construction of the machine learning model involves the use of certain hyperparameters, managed with PyTorch to define the dataset and data loader, trained with Adam's algorithm, and CrossEntropyLoss loss function. The chat was tested with 30 questions from users and the results show that the chatbot can answer user questions quite well with an F1-score of 0.84.