| dc.description.abstract | The question-answer system is now widely used and helps individuals by
automatically responding to frequently asked questions in a variety of industries. These
systems, however, are dependent on the user's context, training data, and the learning
strategies used. As a result, creating such a solid dataset and comprehensive contexts is
required, however it is a difficult task. Through numerous layers of inquiries, Deep Neural
Network may assist in inferring semantic information and offer useful responses to user
enquiries. This thesis suggests a novel Deep Neural Network-based chatbot model to
automatically produce appropriate contextual answers. The suggested chatbot includes a case
study on admissions counseling at International University-Vietnam National University in
Ho Chi Minh City. To see how the system responds to various inquiries, three experiments
were run. The first test looks at the loss experienced when the DNN model is being trained,
the second test is based on a survey and evaluation of nearby users, and the third test is based
on a practical test consisting of 40 questions covering a variety of topics. varying degrees of
difficulty and pass judgment. The main function of the chatbot is to answer questions related
to international university admissions accurately and clearly according to the context,
moreover with the integration of voice asking will make it easier for users. According to test
results, a chatbot powered by Deep Neural Network can provide in-depth responses. The
outcomes are examined to demonstrate the viability and potential of the suggested chatbot. | en_US |
