Dissertation Defense Announcements

Social media platforms such as Facebook, TikTok and Instagram have witnessed increasing use of offensive language by online users, which can be harmful to other users. Recently the continuance of the pandemic has propelled the propagation of offensive content associated with Covid-19 on social media. Some researchers begin to develop effective methods for detecting online offensive language from social media content automatically, yet automatic intervention of offensive language after it is detected remains largely understudied. To address the gaps, this dissertation develops an effective text style transfer-based approach, TST-IOC, for automatic offensive intervention tasks. The promising outcome suggests that our proposed method shows significant potential and could be a preferred choice among users for offensive intervention tasks. This dissertation provides some contributions. First, it contributes significantly to the field of offensive language research by introducing a novel text style transfer-based approach, which has been rarely explored in existing intervention studies. This approach shows a step forward in the development of an automatic offensive intervention system, addressing the limitations of current filtering systems deployed by social media platforms. Second, existing research has mainly focused on using performance metrics for evaluating offensive intervention methods quantitatively. However, this study goes beyond by proposing a comprehensive automatic evaluation paradigm. By exploring both quantitative and qualitative aspects of automatic intervention assessment, it fills a crucial gap in the current offensive language research landscape. Finally, it recognizes the scarcity of studies comparing human evaluation with automatic evaluation in automatic intervention systems. To bridge this gap, we conduct a user study, which allows for an investigation of user acceptance of the proposed automatic intervention approach in real-world scenarios. The insights gained from this user study not only guide the design of more comprehensive automatic intervention systems but also hold the potential to shape the development of human-centric automatic intervention systems in the future.

This dissertation seeks to improve the budget allocation process for economic mobility policy portfolios by leveraging multi-objective optimization as a decision support tool, accounting for population, political, social, and budgetary constraints. Economic mobility is measured as the difference in income between Black and White populations, known as the racial wealth gap. First, I run regression, mediation, and moderated mediation analyses to understand the impact of local authority, consolidation, local partisanship, unified government, and racial demographics on aid, budget expenditures and economic mobility. I then propose a novel application of multi-objective optimization1 to identify optimal mixes aimed at increasing economic mobility in urban cities. In doing so, I seek to improve decision support tools available to local urban governments. My work intends to enable local urban governments to leverage multi-objective optimization to guide their decisions regarding policy selection and budget allocation. Better informed policy processes lead to a better mix of policies, which allows for more holistic solutions with greater societal returns. This not only improves outcomes for residents; it also recovers waste in the governmental process, increasing effectiveness and efficiency.

Beta-lactamase proteins are major contributors to antibiotic resistance, rendering beta-lactam antibiotics ineffective against bacterial infections. The emergence of novel beta-lactamases with expanded substrate specificity poses a global health threat. This study utilizes computational techniques to investigate the mechanisms by which beta-lactamases expand their substrate specificity, enabling bacteria to resist new antibiotics. By exploring the relationship between protein dynamics and function, the impact of enzyme motion on substrate specificity is elucidated.
Molecular dynamics simulations are conducted and analyzed to identify the functional dynamics involved in substrate recognition in beta-lactamase. Dynamic signatures are identified using a novel approach called Supervised Projective Learning with Orthogonal Completeness (SPLOC). Increased flexibility in loops neighboring the enzyme's active site facilitates optimal interactions with different antibiotics through local conformational flexibility. Notably, dynamic signatures differ between protein-antibiotic systems, highlighting the complexity of antibiotic binding mechanisms. These dynamic signatures are demonstrated as viable predictors of antibiotic resistance in beta-lactamase enzymes.
A proof-of-concept is presented for designing de-novo peptides that target these regions, offering a potential new class of beta-lactamase inhibitors capable of hindering the motions necessary for substrate recognition. This approach presents a promising strategy for controlling beta-lactamase-mediated antibiotic resistance.

The increased reliance on Big Data Analytics (BDA) in society, politics, policy, and industry has catalyzed conversations related to the need for promoting ethical reasoning and decision-making in the mathematical sciences. While the majority of professional data scientists today come from privileged positions in society, those processed by the decisions made using data science are more often members of one or more marginalized social groups, translating into disproportionately negative outcomes for these individuals in society. Thus, it is argued that future citizens must develop an ethical mathematics consciousness (EMC) that human beings do mathematics; thus, there are potential ethical dilemmas and implications of mathematical work which may affect entities at the individual, group, societal, and/or environmental level. Drawing from this conjecture, the purpose of this Design-based research study was to develop a local instruction theory and materials that promote students’ ethical mathematics consciousness in a high school Ethical Data Science (EDS) course grounded in a feminist, relational ethic of caring and social response-ability. Outputs include the identification of design heuristics, including the task structures, participation structures, and discursive moves that supported students' development of EMC and equitable participation in classroom activities, an initial curriculum for the EDS course, and a student-use protocol and corresponding analytic framework for making critically conscious ethical decisions in data science.