Eiichiro Kazumori (with collaboration/inputs with UMass Dartmouth students)
April 29, 2024
The bottom line of the state university is education and community. Given the crucial significance of education in the age of AI, it is critical to ensure this bottom line. Nevertheless, currently, students, instructors, and the university inefficiently need to spend a significant amount of time on issues other than actual teaching and learning. Students need to spend a significant amount of time learning about student life, academic resources, technical issues, and campus information by looking at the university webpage, which is not necessarily easy to navigate. Often, students get lost before asking for help. At the same time, instructors and the university need to spend a significant amount of time answering similar questions from students each semester over and over again. This UMass Dartmouth Large Language Model Helper uses LLM to automate the process to reduce these inefficiencies by helping students get answers quickly so that students, instructors, and the university can refocus on their bottom line of education and community.
A use case: a student would want to learn about the University of Massachusetts Dartmouth. A student may type a question: "Tell me about the University of Massachusetts Dartmouth." Then, the chatbot will provide an answer to help students.
Further use cases (not implemented yet): (1) Students, instructors, and staff can check whether a facility is open or not without calling the facility. (2) Students can get course information in the syllabus without contacting instructors each time. (3) Students need not contact the instructor and staff for tech support. (4) Students can learn class offerings that fit their needs and preferences. (5) Students can get their first health support even without going to the health center. (6). Students get community contacts.
The project will save time for students, instructors, and the university. The project will help student recruitment, quality of life, and retention. The project will help students, instructors, and the university to refocus on education and community, the bottom line of the state university.
Disclaimer: This is a proof of concept version. The code has not undergone testing, verification, fine-tuning, and RLHF. This is not a production deployment. It is highly recommended to thoroughly verify the output before proceeding with any actual use.
Caveat: The author would like to thank the students who participated in the initial discussions that inspired this project when Kazumori discussed the generative TA project (see below). Kazumori did all the work on his own. The current presentation of the project is preliminary, ongoing, and is subject to further revision.
We developed the first version in May 2024. The July 2024 version included data from UMass Dartmouth, WHOI, City of New Bedford, Town of Dartmouth, Town of Westport, and Southcoast Health. Due to current resource limitations as we teach classes, we will not have the current version here. Thank you for understanding.
Eiichiro Kazumori
December 29, 2023
Students with disadvantages in social status and economic resources will be put in inherently disadvantageous situations compared with students with social and economic advantages. Ensuring equal opportunities, even for disadvantaged students, regardless of their social and economic status, is one of the most fundamental principles of education, especially in the current unequal society. But, in reality, overcoming such disadvantages is very difficult because disadvantaged students have fewer resources and advantaged students want to maintain their control over disadvantaged students using their resources. The goal of Generative TA is to create practice questions from Youtube videos available for everyone with internet access, even for those who do not have access to human TAs, expensive textbooks, and extensive school resources.
Example: One would want to learn about the production function. One inputs the URL of a YouTube video about the production function (https://www.youtube.com/watch?v=4kFLOoA5l4s) and then inputs the question "What is a production function?" to get a practice question about the concept of the production function.
Browsers: Microsoft Edge and Firefox would work better than Chrome (not fully tested).
Disclaimer: Please note that this server is intended for development purposes only and is not a production deployment. The code provided is a proof-of-concept and has not undergone a rigorous testing process. There are several potential issues, including those related to video quality and server overload, that may affect the performance of the code. It is highly recommended to thoroughly verify the output before proceeding with any actual deployment. If the generated question does not work, please try to generate a new question.
Caveat: This version is publicly available for demos, but the website will soon be password-protected for further development.
We developed the first version in December 2023. Due to current resource limitations as we teach classes, we will not have the current version here. Thank you for understanding.