Telegram FAQ Chatbot Design for Budi Mulia Lawang Junior High School

Abstract

This study effectively created and assessed an academic services chatbot that employs Natural Language Processing (NLP) and Artificial Neural Networks (ANN). It demonstrates a thorough methodology that includes data collection, preprocessing, model training, and performance evaluation. The dataset was meticulously derived from frequently asked questions at the Institute of Business and Technology Asia Malang, ensuring its relevance and applicability to the target audience. Thorough preprocessing methods, such as tokenization, stemming, and the removal of stop words, were utilized to elevate the quality of the input data for the ANN model, leading to a notable enhancement in its performance. The training outcomes demonstrated a significant relationship between the number of training epochs and the accuracy of the chatbot's responses, suggesting that more extensive training resulted in improved performance. Additionally, the measurement tool employed for user feedback demonstrated confirmed validity and reliability, evidenced by a Cronbach's Alpha coefficient of 0.965, highlighting the strength of the data gathered. A study conducted with 37 students revealed a significant level of satisfaction regarding the chatbot's performance, which attained an impeccable accuracy score of 100%. The results underscore the promise of chatbots powered by natural language processing to improve academic information services, efficiently responding to student questions and markedly alleviating the burden on academic personnel. This study provides a significant framework for educational institutions looking to adopt AI-driven solutions to enhance their academic support services and enrich the student experience.