Virtual Assistant Prototype to Answer Student Payment and Enrollment Questions Through Natural Language Processing and Machine Learning
Franklin Parrales–Bravo, Rosangela Caicedo–Quiroz, Julio Barzola–Monteses, José Guillén-Mirabá, Otto Guzmán-Bedor · 2024
The application proposed in this research is aimed at students of the Software Engineering degree at the Faculty of Mathematical and Physical Sciences, who have problems accessing information about tuition payments and enrollment. For this purpose, a virtual assistant is proposed that can resolve the doubts of the students of the software career at the University of Guayaquil, using Natural Language Processing (NLP) and Machine Learning (ML). The virtual assistant is integrated with the Telegram messaging platform, allowing students to access the information they need at any time and anywhere, without visiting the faculty in person. To evaluate the user experience when using the proposed virtual assistant, a survey was conducted with 60 students who previously interacted with the proposed application. Aspects such as Ease of use of the system (A), speed (B), appropriate behavior (C), confidence to use it again (D), and quality of the answer provided (E) were evaluated. All these items were scored using a Likert scale whose values are represented by the numbers from 1 to 5, in that order: extremely dissatisfied, disappointed, acceptable, adequate, and excellent. Each evaluated characteristic (A-E) was mostly rated with a grade greater than 3. Therefore, the proposed virtual assistant has achieved a good perception of the use of the system, even considering it for use in future administrative processes of students in the faculty.