Date of Graduation
12-2024
Document Type
Thesis
Degree Name
Master of Science in Computer Science (MS)
Degree Level
Graduate
Department
Electrical Engineering and Computer Science
Advisor/Mentor
Gauch, Susan E.
Committee Member
Panda, Brajendra N.
Second Committee Member
Zhang, Lu
Keywords
Artificial Intelligence; Diffusion; Generative AI; Recommendation System
Abstract
This thesis addresses the problem of recommending items to users based on their ratings of items through a diffusion recommender system. Regular recommender systems are already capable of efficient recommendation through conventional methods such as collaborative or content-based filtering. Diffusion is a new type of generative AI that aims to improve our previous AI's shortcomings in the generative domain, like Variational Autoencoders (VAEs) and Generative Adversarial Networks (GANs). We use diffusion to create a recommender system that mirrors the sequence users take when browsing and rating items. The current methods of recommendation with diffusion do not use the new innovation in diffusion models as a whole, known as classifier-free guidance. In this thesis, we aim to improve the previous approaches of diffusion recommender systems by implementing classifier-free guidance and augmenting the underlying model that powers diffusion recommender systems. We implement several different architectures for the underlying model and many different methods, including classifier-free guidance. Our findings show improvements to a previous implementation for most metrics on most datasets and an enhanced ability to create relevant recommendations for users from sparse and not information-dense datasets, where there are far fewer items than users.
Citation
Buchanan, N. (2024). Exploring the Capabilities of Classifier-Free Guidance in Recommendation Tasks. Graduate Theses and Dissertations Retrieved from https://scholarworks.uark.edu/etd/5538
