DIGITAL LIBRARY
CHATBOT TO SUPPORT STUDENTS IN DEVELOPING A TERMINAL PROJECT PROPOSAL
Universidad Autónoma Metropolitana (MEXICO)
About this paper:
Appears in: EDULEARN24 Proceedings
Publication year: 2024
Pages: 6472-6479
ISBN: 978-84-09-62938-1
ISSN: 2340-1117
doi: 10.21125/edulearn.2024.1530
Conference name: 16th International Conference on Education and New Learning Technologies
Dates: 1-3 July, 2024
Location: Palma, Spain
Abstract:
To successfully complete the Computer Engineering program at Universidad Autónoma Metropolitana in Mexico City, students can develop a technology or research project under the supervision of a tutor. However, they must present and to approve a formal document, named project proposal, describing the project. At least two members of the Computing Engineering Committee review and authorize the project proposal. In case of approval, the next quarter, students develop the project.

The project proposal is considered a very important step in professional development because students apply their knowledge and skills in problem-solving like Computer Engineering professionals.

Committee members evaluate each project proposal with a framework of 14 sections and one final evaluation. After that, the Program Coordinator gives a qualitative and qualitative evaluation that is sent to the student and his/her tutor.

During the first weeks of the quarter, the Program Coordinator guides the student in the writing of each section with the support of his/her tutor. Before the student submits his/her work, it is very convenient to do an auto-evaluation that reduces errors and possible omissions. To support this auto-evaluation, a chatbot available through a web app was created to allow those who wish to receive feedback.

The chatbot was created in Python using ChatGPT and Streamlit. An iterative methodology (idea, implementation, and experimental result) was followed for the creation of the following prompts:
- Review that all the sections are present and in the order specified.
- Determine the 5 main topics.
- Check spelling.
- Proofreading and editing of the text.
- References are in IEEE format.

Chatbot evaluation was performed using 20 past project proposals using the ChatGPT web interface for each prompt by at least two Committee members comparing the results generated by the system and the human evaluation. Results are promising, so future work includes delivering the chatbot results to the students.
Keywords:
Chatbot, ChatGPT, problem-solving skill, educational aids.