The need to form teams happens all the time, but different people want to form teams in different ways. We have developed a plugin to Canvas that forms teams using some algorithms, but more work is needed to provide alternative algorithms. The plugin also needs a peer evaluation component as well as a visual dashboard implemented. For integration compliance reasons, this project is built using Python/Django.
A web platform that helps students practice introductory programming. We have done a lot of work to develop novel content and automated testing, but there's always room for more work. The website has various features for student learners as well as the administrator. More work is needed to develop the navigation front end (most likely Angular), the user dashboards, and some AI in the backend to monitor student usage in real time. For integration compliance reasons, this project is built using Python/Django.
Computational thinking is a recent buzz term that is crucial in problem solving, scientific inquiry, and coding. A lot of debate in the recent literature has tried to define precisely what this term entails, and how that may be assessed in schools, or how it may impact teaching and student learning. An interesting aspect of this work is the development and evaluation of non-screen resources and/or commercial products that can introduce skills relevant to coding for little kids.
Analyzing whether students develop the right skills for the job market is important for educators. I have a dataset to analyze student skills and to compare them with job postings and educational programs. Exploration in this project will involve data analysis, web scraping, natural language processing, statistical analyses, and research into relevant literature on the topic. You may also do comparisons across subject domains, such as computer science versus marketing.
In an ideal world, students should get immediate feedback on their submitted work. In an effort to accomplish this, I am exploring several methods for automating feedback. Due to the nature of unstructured code, machine learning methods such as unsupervised learning and semi-supervised learning are needed. This project requires a strong background statistics and a good handle on common machine learning techniques.
There are lots of mobile educational games around but not much about how they should be designed. The literature also lacks assessments in these games. This project involves the analysis of existing mobile educational games and related literature on their development and assessments. You will need to identify specific age groups (or specific educational levels such as concepts for grade 6 curriculum), map onto cognitive skills required to excel in these games and to master the intended skills, assess chance of success if user just randomly clicks, document measure of support the games have, etc. You will also want to consider phone versus tablet versus desktop interfaces, since they all come with different limitations.
Text analysis and autograding is a hot area that many educators in AI are exploring. This project involves the use of various natural language processing and machine learning techniques. The goal here is to use past data to develop a model that can categorize unseen text with the appropriate labels. The focus is on interpretable models, but an evaluation using other methods will be needed as well.