Peran AI Chatbots dalam Layanan Mahasiswa Ilmu Komunikasi Menggunakan Metode Natural Languange Processing (NLP) di Universitas Muhammadiyah Karanganyar

Abstract

Three technical steps of data collection were used in testing that involved over 100 students: observations, questionnaires, and interviews. After processing, all of the data is collected into a dataset that is kept in the chatbot application database. Obtaining up-to-date information about class schedules, communication science resources, and other academic matters is crucial for students. Therefore, this project intends to provide an accurate and up-to-date source of information by using AI chatbots with the NLP approach, making it easier for students to acquire the material they need. In the exam scenario, each student is required to respond to a minimum of five questions that follow the topics covered in Communication Science curriculum. These questions are processed using the application provided to produce predictions of possible relevant answers. The results of the answers from the chatbot are used to create a comparison matrix between the percentage of chatbot usage and the verification team. The verification team approves all output from the chatbot. Based on the comparison, it can be inferred that the verification team's output percentage was 100%, but the chatbot's production percentage was 95%.