UROP Project

HOT-KG: Harmonic Optimal Transport for Knowledge Graph Completion

machine learning; artificial intelligence; knowledge graphs; mathematics; link prediction
Research Mentor: Astrit Tola, Dr.
Department, College, Affiliation: Mathematics, Arts and Sciences
Contact Email: atola@fsu.edu
Research Assistant Supervisor (if different from mentor):
Research Assistant Supervisor Email:
Faculty Collaborators:
Faculty Collaborators Email:
Looking for Research Assistants: Yes
Number of Research Assistants: 1
Relevant Majors: Open to students of all majors.
Students from the following programs may find the material connects especially well to their coursework: Computer Science; Mathematics; Statistics; Data Science; Scientific Computing; Computational Science; Physics; Electrical and Computer Engineering; Information Technology.
Project Location: On FSU Main Campus
Research Assistant Transportation Required:
Remote or In-person: Partially Remote
Approximate Weekly Hours: 10, Flexible schedule (Combination of business and outside of business. TBD between student and research mentor.)
Roundtable Times and Zoom Link:
Not participating in the roundtable

Project Description

Almost every AI system you use every day depends on a knowledge graph, a huge collection of facts written as simple statements: Tallahassee - is the capital of - Florida. When you ask a voice assistant a question, when a search engine shows you a box of facts about a person, when a streaming service explains why it recommended something, a knowledge graph is doing the work behind the scenes. Hospitals and pharmaceutical researchers use them to connect drugs, genes, and diseases; banks use them to trace patterns of fraud.

The problem is that these graphs are always incomplete. Even the largest ones are missing an enormous number of true facts, simply because nobody ever wrote them down. So a central question in machine learning is: given the facts we already have, can a computer figure out which facts are missing? The goal of this project is to design and test a new approach to knowledge-graph link prediction. We will investigate whether the proposed model can identify missing facts more accurately than current methods, particularly for relations with multiple possible answers.



Research Tasks: The student will review recent knowledge-graph completion research from major AI conferences, maintain a shared reference library, and prepare short, plain-language summaries for discussion and use in the paper’s background section. They will also locate and reproduce public implementations of several established methods, troubleshoot the code as needed, and verify reported results. With my guidance and a provided mathematical specification, the student will implement and test the project’s core scoring function in Python using PyTorch and an optimal-transport library.

Skills that research assistant(s) may need: Recommended - Basic programming in Python

Mentoring Philosophy

My mentoring philosophy is centered on meeting students where they are and providing the support, instruction, and encouragement they need to grow. I begin by learning about each student’s goals, interests, strengths, and current level of preparation. I identify gaps in their knowledge and directly teach the concepts or skills they need to move forward, adjusting my explanations and expectations so that the material remains challenging but understandable. I believe effective mentoring should build on what a student already does well while helping them develop new abilities. Together, we break larger projects into clear and manageable milestones, regularly review progress, and address difficulties before they become barriers. I provide honest and constructive feedback by clearly identifying areas that need improvement and offering specific guidance on how the student can improve. As students gain knowledge and confidence, I gradually reduce my level of guidance and encourage them to take greater ownership of their work. I ask students to investigate questions, compare possible approaches, explain their reasoning, and defend their conclusions. This process helps them develop independence while remaining accountable for their preparation, decisions, and progress.

Additional Information


Link to Publications

https://astrittola.github.io/publications.html