UC computer science engineer works to improve AI explainability

I have always wanted to research the areas of databases and machine learning within computer science. When I was applying for graduate programs, the University of Cincinnati stood out to me due to its status as an R1 research institution and its cutting-edge initiatives, strong academic programs, renowned faculty, vibrant campus communities and campus resources. I was also interested in the research of Seokki Lee, assistant professor of computer science. His work centers on databases, data provenance, and machine learning, or ML, which perfectly aligns with my interests. These factors made UC the ideal choice for my graduate journey. 

My research focuses on developing explainable ML models using data provenance, or data lineage. Many industries and sectors rely heavily on ML and AI algorithms for data-driven decision-making, which enhances operational efficiency. However, there is often a general mistrust of their predictions and outputs, primarily due to their lack of explainability as black-box algorithms (algorithms that do not allow users to see the logic that produced the output). For example, consider a scenario where an AI model predicts that a credit card applicant will default on payments and denies their credit application, but does not give clear insights as to why. Or, imagine a situation wherein a facial recognition algorithm produces erroneous outputs leading to unjust accusations or arrests. To address this challenge and build trust in ML predictions and outputs, explainability is crucial. 

Through my research work, I am committed to developing provenance-based explainable solutions that enhance ML explainability, mitigate biases, and increase adoption of this technology. My objective is to advance ML explainability by leveraging data provenance to provide human-understandable explanations for ML predictions and outputs. This interdisciplinary research combines databases, data provenance, and ML within the Database Group at UC led by Dr. Seokki Lee. 

I aim to bridge the gap between the technical complexities of AI/ML and users with diverse expertise, ensuring transparency and trustworthiness in ML applications. 

The personal and professional growth I have experienced during my time at UC has been remarkable. One achievement I'm proud of is my research project on developing provenance-based solutions for explainable machine learning models. I have presented this work at prestigious conferences like the Institute of Electrical and Electronics Engineers International Conference on Data Engineering, American Association for the Advancement of Science Annual Meeting, and the Greater Chicago Area Systems Research Workshop.

In addition, I have had the honor of contributing to the scholarly community by serving as a program committee member and a reviewer on the research paper track for various conferences. 

I take great pride in my contributions to professional organizations in science and engineering such as the Institute of Electrical and Electronics Engineers. Currently, I am serving a two-year leadership commitment on the Society of Women Engineers' national organization and was honored to speak at two panel sessions dedicated to advancing graduate engineering at the 2023 conference. 

Equally fulfilling has been my role in nurturing the next generation of scientists and engineers. I served as a mentor for the UC McNair Scholars, the Society of Women Engineers' "Invent It, Build It" program for K-12 students, and as a judge at events such as the Ohio Academy of Science District Science Day and State Science Day. I was also a counselor at various CEAS summer camps and a professional reviewer of research posters at the 2023 UC Undergraduate Scholarly Showcase. 

Reflecting on my journey, I am grateful for the consistency and excellence that have marked my experience. Earning my master's degree in 2023 from UC was a significant milestone towards completing my doctorate. I am also honored by the accolades I have received, including being named Graduate Student Engineer of the Month, and receiving the 2023 Cadence Diversity in Technology Scholarship Award. 

I plan to defend my PhD dissertation by spring 2026. After completing my PhD, I will pursue a career in research or industry, specializing in developing optimized, large-scale unbiased data provenance systems and ML solutions. My objective is to contribute to pioneering scientific advancements, ideally in an academic or high-tech industry environment. 

Feature image at top: Nkechi Jennifer Akinwale hopes to improve trust and explainability of AI and Machine Learning. Photo/unsplash