UMBC Faculty Collection
Permanent URI for this collectionhttp://hdl.handle.net/11603/15
Browse
Recent Submissions
Item ADAPTIVE IE: Investigating the Complementarity of Human-AI Collaboration to Adaptively Extract Information on-the-fly(ACL, 2025-01) Mondal, Ishani; Yuan, Michelle; N, Anandhavelu; Garimella, Aparna; Ferraro, Francis; Blair-Stanek, Andrew; Van Durme, Benjamin; Boyd-Graber, JordanInformation extraction (IE) needs vary over time, where a flexible information extraction (IE) system can be useful. Despite this, existing IE systems are either fully supervised, requiring expensive human annotations, or fully unsupervised, extracting information that often do not cater to user`s needs. To address these issues, we formally introduce the task of “IE on-the-fly”, and address the problem using our proposed Adaptive IE framework that uses human-in-the-loop refinement to adapt to changing user questions. Through human experiments on three diverse datasets, we demonstrate that Adaptive IE is a domain-agnostic, responsive, efficient framework for helping users access useful information while quickly reorganizing information in response to evolving information needs.Item Can Generative AI be Egalitarian?(IEEE, 2024-10) Feldman, Philip; Foulds, James; Pan, ShimeiThe recent explosion of “foundation” generative AI models has been built upon the extensive extraction of value from online sources, often without corresponding reciprocation. This pattern mirrors and intensifies the extractive practices of surveillance capitalism [46], while the potential for enormous profit has challenged technology organizations’ commitments to responsible AI practices, raising significant ethical and societal concerns. However, a promising alternative is emerging: the development of models that rely on content willingly and collaboratively provided by users. This article explores this “egalitarian” approach to generative AI, taking inspiration from the successful model of Wikipedia. We explore the potential implications of this approach for the design, development, and constraints of future foundation models. We argue that such an approach is not only ethically sound but may also lead to models that are more responsive to user needs, more diverse in their training data, and ultimately more aligned with societal values. Furthermore, we explore potential challenges and limitations of this approach, including issues of scalability, quality control, and potential biases inherent in volunteercontributed content.Item Fair Inference for Discrete Latent Variable Models: An Intersectional Approach(ACM, 2024-09-04) Islam, Rashidul; Pan, Shimei; Foulds, JamesIt is now widely acknowledged that machine learning models, trained on data without due care, often exhibit discriminatory behavior. Traditional fairness research has mainly focused on supervised learning tasks, particularly classification. While fairness in unsupervised learning has received some attention, the literature has primarily addressed fair representation learning of continuous embeddings. This paper, however, takes a different approach by investigating fairness in unsupervised learning using graphical models with discrete latent variables. We develop a fair stochastic variational inference method for discrete latent variables. Our approach uses a fairness penalty on the variational distribution that reflects the principles of intersectionality, a comprehensive perspective on fairness from the fields of law, social sciences, and humanities. Intersectional fairness brings the challenge of data sparsity in minibatches, which we address via a stochastic approximation approach. We first show the utility of our method in improving equity and fairness for clustering using naïve Bayes and Gaussian mixture models on benchmark datasets. To demonstrate the generality of our approach and its potential for real-world impact, we then develop a specialized graphical model for criminal justice risk assessments, and use our fairness approach to prevent the inferences from encoding unfair societal biases.Item ANSR-DT: An Adaptive Neuro-Symbolic Learning and Reasoning Framework for Digital Twins(2025-01-15) Hakim, Safayat Bin; Adil, Muhammad; Velasquez, Alvaro; Song, HoubingIn this paper, we propose an Adaptive Neuro-Symbolic Learning Framework for digital twin technology called ``ANSR-DT." Our approach combines pattern recognition algorithms with reinforcement learning and symbolic reasoning to enable real-time learning and adaptive intelligence. This integration enhances the understanding of the environment and promotes continuous learning, leading to better and more effective decision-making in real-time for applications that require human-machine collaboration. We evaluated the \textit{ANSR-DT} framework for its ability to learn and adapt to dynamic patterns, observing significant improvements in decision accuracy, reliability, and interpretability when compared to existing state-of-the-art methods. However, challenges still exist in extracting and integrating symbolic rules in complex environments, which limits the full potential of our framework in heterogeneous settings. Moreover, our ongoing research aims to address this issue in the future by ensuring seamless integration of neural models at large. In addition, our open-source implementation promotes reproducibility and encourages future research to build on our foundational work.Item Assessing the K₂BO₃ family of materials as multiferroics(APS, 2024-11-26) Casale, Anthony; Bennett, JosephWe evaluate the potential of an overlooked family of materials to support both the magnetization and polarization required to be classified as multiferroics. This family of materials has a stoichiometry of A₂BX₃ and was uncovered in the Inorganic Crystal Structure Database (ICSD) while searching for structural platforms that could support low energy polarization switching. The examples here have the general chemical formula of K₂BO₃, where B is a magnetically active cation located within edge-sharing square pyramids that form a 1D chain. Density functional theory with Hubbard U corrections (DFT + U) are used to determine the potential energy landscape of K₂BO₃, which include investigating multiple magnetic and polarization orderings. We analyze the ground state and electronic structures and report on how the choice of Hubbard U will affect both, which is important when predicting functional properties of low-dimensional and potentially exfoliable systems. This family contains a ferromagnetic insulator, K₂VO₃, as well as antiferromagnetic (K₂NbO₃) and nonmagnetic (K₂MoO₃) insulators with antipolar ground state symmetries, and accessible polar metastable states, that we predict to be antiferroelectric. This preliminary assessment of the K₂BO₃ members of the A₂BX₃ family reveals a new class of materials, that with further optimization via compositional tuning, could be multiferroic.Item Cultural foundations of global health: a critical examination of universal child feeding recommendations(BMC, 2025-01-23) Scheidecker, Gabriel; Funk, Leberecht; Chaudhary, Nandita; Chapin, Bambi L.; Schmidt, Wiebke J.; El Ouardani, ChristineThere has been a rising call to decolonize global health so that it more fully includes the concerns, knowledge, and research from people all over the world. This endeavor can only succeed, we argue, if we also recognize that much of established global health doctrine is rooted in Euro-American beliefs, values, and practice rather than being culturally neutral. This paper examines the cultural biases of child feeding recommendations as a case in point. We argue that the global promotion of Responsive Feeding—a set of allegedly best practices for child feeding promulgated by the WHO and others—is based on a tacit conviction that certain Western middle-class feeding practices are universally best, along with a promise that future evidence will demonstrate their superiority. These recommendations denounce feeding practices that diverge from this style as Non-Responsive Feeding, thereby pathologizing the many valued ways of feeding children in communities all over the world without sound scientific evidence. Drawing on ethnographic research, we show that there is a wide variety in feeding practices around the world and these are closely interlinked with the understandings and priorities of caregivers, as well as with favored forms of relationships and ways of maintaining them. For global health nutrition interventions to be justified and effective, they would need to be based on more pertinent, culturally responsive research than they currently are. We suggest the use of ethnographic research as an important tool in building empirically grounded, epistemically inclusive, and locally meaningful approaches to improving nutritional support for children in communities around the world and to global health efforts more broadly.Item Digital skills use profiles among older workers in the United States: a person-centered approach(Taylor & Francis, 2024-12-22) Yamashita, Takashi; Narine, Donnette; Ojomo, Adeola; Chidebe, Runcie C. W.; Cummins, Phyllis A.; Kramer, Jenna W.; Karam, Rita; Smith, Thomas J.Considering the digitalisation of the workplace and increasingly crucial digital skill proficiency in the technology-rich labour market, the objectives of the present study are to develop digital skill use profiles and to identify specific individual characteristics that are linked with digital skill use patterns among older workers in the United States. However, relatively little is known about older workers’ digital skill use patterns and skill use opportunity structures. Data of the U.S. older workers (age 50 years and older; n = 1,670) were obtained from the 2012/2014/2017 International Assessment of the Adult Competencies (PIAAC). Latent class analysis – a form of person-centred approach that identifies subgroups based on distinctive digital skill use patterns, showed that there were two underlying subgroups of older workers, including more frequent and less frequent digital skill users. More frequent users practiced a greater variety of digital skills both at work and outside of work than their counterparts. Also, logistic regression analysis showed that higher digital skill proficiency and full-time employment (vs. part-time) were associated with belonging to the more frequent digital skill use subgroup. The digital skill use profiles of U.S. older workers, subgroup characteristics, and implications for adult education and labour policies are evaluated.Item What $2bn in Spending Cuts Looks Like(I Hate Politics Podcast, 2025-01-24) Dasgupta, Sunil; Moon, David; Shetty, EmilyFacing a fiscal crisis, MD Governor Wes Moore has proposed $2 billion dollars in spending cuts in his 2025 budget proposal (as well as taxes and fees). What does that $2 billion look like to legislators beginning to hear from constituents and advocates? Sunil Dasgupta asks Maryland House Majority Leader David Moon and the Appropriations Committee member Emily Shetty to break down the cuts and put them into perspective. Newly in public domain music by Clara Smith and The Troubadours.Item I Hate the News Jan 21(I Hate Politics Podcast, 2025-01-21) Dasgupta, SunilThe weekly news analysis from I Hate Politics: Donald Trump is inaugurated as the 47th president of the US amidst fears in blue states like Maryland where hundreds of thousands of federal workers and contractors live and work. Maryland Governor Wes Moore presents a balanced budget for 2025 with $1.4 bn in cuts, $1.3 bn new revenue, and $800 m in transfers from savings. Newly in public domain music from the 1920s: The Benson Orchestra of Chicago, the Paul Whiteman band, Carl Fenton, and Jan Garber.Item I Hate the News Jan 14(I Hate Politics Podcast, 2025-01-14) Dasgupta, SunilThe weekly news analysis from I Hate Politics: Montgomery County Councilmember Will Jawando makes a big political play on housing policy. State could cut 5 percent from the University System of Maryland budget. The Town of Cheverly in Prince George’s County files lawsuits against the neighboring Bladensburg for trying to annex land on which a big development project is planned. Newly in public domain music from the 1920s: The Benson Orchestra of Chicago, the Paul Whiteman band, Carl Fenton, and Jan Garber.Item What to do About Continuing Racism in Schools(I Hate Politics Podcast, 2025-01-17) Dasgupta, Sunil; Khademian, Leila; Taiwo, TeemoMCPS Student Takeover: Lorena Treviño of Walter Johnson High School talks with Leila Khademian and Teemo Taiwo, co-presidents of Wootton High School’s Black Student Union, about continuing racist incidents and adult failures in one of the academically best-performing high schools in Montgomery County, MD. Newly in public domain music by Nacio Herb Brown and the Carl Fenton band.Item Understanding the Challenges of Maker Entrepreneurship(ACM, 2025-01-23) Friedman, Natalie; Bremers, Alexandra; Nyanyo, Adelaide; Clark, Ian; Kotturi, Yasmine; Dabbish, Laura; Ju, Wendy; Martelaro, NikolasThe maker movement embodies a resurgence in DIY creation, merging physical craftsmanship and arts with digital technology support. However, mere technological skills and creativity are insufficient for economically and psychologically sustainable practice. By illuminating and smoothing the path from ``maker" to ``maker entrepreneur," we can help broaden the viability of making as a livelihood. Our research centers on makers who design, produce, and sell physical goods. In this work, we explore the transition to entrepreneurship for these makers and how technology can facilitate this transition online and offline. We present results from interviews with 20 USA-based maker entrepreneurs {(i.e., lamps, stickers)}, six creative service entrepreneurs {(i.e., photographers, fabrication)}, and seven support personnel (i.e., art curator, incubator director). Our findings reveal that many maker entrepreneurs 1) are makers first and entrepreneurs second; 2) struggle with business logistics and learn business skills as they go; and 3) are motivated by non-monetary values. We discuss training and technology-based design implications and opportunities for addressing challenges in developing economically sustainable businesses around making.Item Review of The Objects of Credence(2025-01-22) Cassell, LisaItem RAGged Edges: The Double-Edged Sword of Retrieval-Augmented Chatbots(2024-06-12) Feldman, Philip; Foulds, James; Pan, ShimeiLarge language models (LLMs) like ChatGPT demonstrate the remarkable progress of artificial intelligence. However, their tendency to hallucinate -- generate plausible but false information -- poses a significant challenge. This issue is critical, as seen in recent court cases where ChatGPT's use led to citations of non-existent legal rulings. This paper explores how Retrieval-Augmented Generation (RAG) can counter hallucinations by integrating external knowledge with prompts. We empirically evaluate RAG against standard LLMs using prompts designed to induce hallucinations. Our results show that RAG increases accuracy in some cases, but can still be misled when prompts directly contradict the model's pre-trained understanding. These findings highlight the complex nature of hallucinations and the need for more robust solutions to ensure LLM reliability in real-world applications. We offer practical recommendations for RAG deployment and discuss implications for the development of more trustworthy LLMs.Item Mononuclear Aluminum–Fluoride Ions, AlFx(⁺/⁻)—Study of Plausible Frameworks of Complexes with Biomolecules and Their In Vitro Toxicity(MDPI, 2025-1) Pavlovič, Anja; Janžič, Larisa; Sršen, Lucija; Kopitar, Andreja Nataša; Edwards, Kathleen F.; Liebman, Joel F.; Ponikvar-Svet, MajaThe importance of fluorine and aluminum in all aspects of daily life has led to an enormous increase in human exposure to these elements in their various forms. It is therefore important to understand the routes of exposure and to investigate and understand the potential toxicity. Of particular concern are aluminum–fluoride complexes (AlFx), which are able to mimic the natural isostructural phosphate group and influence the activity of numerous essential phosphoryl transferases. Our review of salts of ionic AlFx species, which plausibly form the framework of complexes with biomolecules, revealed that the octahedral configuration of aluminum in the active site of the enzyme is preferred over the trigonal-bipyramidal structure. The effects of varying concentrations of fluoride, aluminum and AlFx—from micromolar to millimolar levels—on the viability and apoptosis rate of THP-1 monocytes were investigated using phosphate buffer solution as a culture media to simulate physiological conditions. Our results suggest that aluminum can reduce the direct toxicity of fluoride through the formation of AlFx. In view of the results found, further in vitro studies are required to clarify the toxicity mechanisms of these species.Item GenderAlign: An Alignment Dataset for Mitigating Gender Bias in Large Language Models(2024-12-16) Zhang, Tao; Zeng, Ziqian; Xiao, Yuxiang; Zhuang, Huiping; Chen, Cen; Foulds, James; Pan, ShimeiLarge Language Models (LLMs) are prone to generating content that exhibits gender biases, raising significant ethical concerns. Alignment, the process of fine-tuning LLMs to better align with desired behaviors, is recognized as an effective approach to mitigate gender biases. Although proprietary LLMs have made significant strides in mitigating gender bias, their alignment datasets are not publicly available. The commonly used and publicly available alignment dataset, HH-RLHF, still exhibits gender bias to some extent. There is a lack of publicly available alignment datasets specifically designed to address gender bias. Hence, we developed a new dataset named GenderAlign, aiming at mitigating a comprehensive set of gender biases in LLMs. This dataset comprises 8k single-turn dialogues, each paired with a "chosen" and a "rejected" response. Compared to the "rejected" responses, the "chosen" responses demonstrate lower levels of gender bias and higher quality. Furthermore, we categorized the gender biases in the "rejected" responses of GenderAlign into 4 principal categories. The experimental results show the effectiveness of GenderAlign in reducing gender bias in LLMs.Item Describing seasonal mixtures of cloud regimes via “regimes of regimes”(AMS, 2025-01-22) Cho, Nayeong; Oreopoulos, Lazaros; Lee, Dongmin; Tan, Jackson; Jin, DaehoWe propose a new type of cloud classification, relevant to monthly or longer time scales, but which inherently still encompasses cloud subgrid variability information at ~100 km scales. Our proposed classification partitions frequencies of occurrence over these scales of previously defined cloud regimes (CRs). We call the resulting distinct cloud entities regimes of regimes (RORs). While the CRs have been previously shown to successfully classify daily mesoscale subgrid variability via distributions of cloud fraction within distinct combinations of cloud top pressure and cloud optical thickness, the RORs essentially represent the prevalent seasonal mixtures of these CRs. RORs thus embody the seasonal cloudiness of a mesoscale region. We show that each ROR can still be associated with more traditional cloud classifications via composites of coincident active (lidar and cloud radar) cloud views. In a first application that gauges the potential utility of RORs, we pair them with CERES EBAF radiative fluxes to gain insight into recent trends of the cloud radiative effect. The ROR corresponding to an environment of shallow convection stands out in this analysis largely because of its declining population. Our study demonstrates the potential of RORs to categorize globally mesoscale cloudiness at monthly/seasonal scales and to serve as proxies of different atmospheric states at these scales.Item Advancing Ensemble Streamflow Prediction Through Satellite-based Precipitation Product and Model Parameter Uncertainty Quantification(2025-1-18) Peng, Kaidi; Wright, Daniel Benjamin; Derin, Yagmur; Alexander, Gary Aaron; Pradhan, Ankita; Zoccatelli, Davide; Hartke, Samantha H.; Li, Zhe; Tan, JacksonSatellite-based quantitative precipitation estimates (QPE), such as NASA’s Integrated Multi-satellitE Retrievals for GPM (IMERG), provide easily accessible continental-to-global precipitation forcings for flood prediction and other hydrologic applications. Nevertheless, when used in hydrologic prediction, uncertainty in satellite-based QPE often leads to significant bias. This forcing uncertainty is further blended with other error sources, including process representation, parameter values, and their interactions. The identification and decoupling of these uncertainties can enhance our understanding of their respective impacts, thereby improving hydrologic prediction. Addressing this issue worldwide is challenging, however, largely due to the scarcity of precipitation ground truth and complex uncertainty interactions. Therefore, we propose an efficient uncertainty quantification framework for ensemble streamflow prediction, which keeps different uncertainty sources separable through hierarchical Bayesian inference. Satellite-based QPE uncertainty is characterized by a novel near-realtime quasi-global satellite-only ensemble precipitation dataset (STREAM-Sat), which is completely independent of ground-based precipitation measurements. Model parameter uncertainty in a distributed physics-based hydrologic model is inferred by an Iterative Ensemble Smoother (IES). To illustrate the impact and limitations of precipitation uncertainty, we compared ensemble streamflow predictions driven by both model parameter and satellite precipitation uncertainties and ensemble streamflow predictions driven by model parameter uncertainty and deterministic QPE. We demonstrate that the quantification of satellite-based QPE uncertainty notably improves the accuracy and reliability of streamflow predictions. This study also lays a foundation for satellite-based streamflow prediction in ungauged regions.Item Topological X-states in a quantum impurity model(2025-01-26) Cavalcante, M. F.; Bonança, Marcus V. S.; Miranda, Eduardo; Deffner, SebastianTopological qubits are inherently resistant to noise and errors. However, experimental demonstrations have been elusive as their realization and control is highly complex. In the present work, we demonstrate the emergence of topological X-states in the long-time response of a locally perturbed quantum impurity model. The emergence of the double-qubit state is heralded by the lack of decay of the response function as well as the out-of-time order correlator signifying the trapping of excitations, and hence information in local edge modes.Item The Uncertain Future of I-O Licensing and Certification: The SIOP Certification Task Force Requests Your Attention(SIOP, 2023) Shoenfelt, Elizabeth L.; Lasson, Elliot; Lefkowitz, Joel; Lewis, Robert E.; Lowman, Rodney L.; Schroeder, Daniel A.; Walters, Judith C.