THE ARTIFICIAL INTELLIGENCE CHATBOT EDUCATIONAL CONSULTING: A CASE STUDY OF THE FACULTY STUDENTS SRIPATUM UNIVERSITY

Abstract

The objectives of this research are (1) to develop an intelligent chatbot system for academic
consultation and to enhance machine learning capabilities (2) to evaluate the effectiveness of
the intelligent chatbot system and (3) to assess user satisfaction with the intelligent chatbot system.
In this research, the development of the intelligent chatbot system is carried out through LINE
programming connected to a question-answer database in Google Sheets. Using Jason and Python
scripting app languages, the system learns questions using a Sequence to Sequence Model, focusing
on specific parts of the question sentence that are of interest for learning. The system then provides
responses accordingly. The performance evaluation of the system by three experts found that the overall
assessment was at a high level (x = 4.44, SD = 0.51). Additionally, the satisfaction assessment from
a purposive sample of 200 participants, including 10 faculty or staff members and 190 students, revealed
that overall satisfaction was at a high level (x= 4.04, SD = 0.85). The findings of this research include
the utilization of a selective learning machine model focusing on parts of question sentences of interest,
learned from a dataset of 500 question-answer pairs. This has resulted in increased efficiency and reduced
response time for consultation services provided by center staff.