EVALUASI KINERJA CHATBOT DENGAN INTEGRASI ALGORITMA RNN DAN LSTM DALAM OPTIMALISASI RESPON PERCAKAPAN PADA SISTEM PMB

Abstract

AI-based chatbots are increasingly used in various sectors to improve service efficiency, including in education. This research evaluates the performance of a chatbot on the New Student Admission (PMB) system with the integration of Recurrent Neural Network (RNN) and Long Short-Term Memory (LSTM) algorithms. In its implementation, chatbots play an important role in providing information and answering user questions automatically, but often have difficulty in understanding complex conversational contexts. The use of RNN and LSTM algorithms is expected to overcome the limitations of traditional chatbots in understanding conversational context and providing more relevant responses. The evaluation results show that the integration of RNN and LSTM is able to improve the quality of chatbot responses, both in terms of accuracy of 92.86% and relevance in complex conversation scenarios. The proposed chatbot proved to be effective in understanding user requests and providing faster answer responses compared to the conventional methods used. This implementation provides a more optimal solution in the PMB system, which is expected to be implemented at STMIK AKI Pati.