EMEmma Louise McDaniel

Emma Louise McDaniel

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Emma Louise McDaniel has a total of 25 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Emulating Hydrodynamic Models with Neural Operators for Rapid Storm Assessment

    by Emma L. McDaniel, Jeffrey Liu, Chad Council, Jenny Rowlett, Teresa Fazio, Julia Hopkins

    Timely, accurate, and trustworthy flood models for rapid decision making during and after extreme storms do not exist in an operational capacity, as flood models often compromise between physical accuracy and speed. Physics-based flood models that explicitly simulate mass and momentum balances can be accurate, but are prohibitively computationally expensive. Stochastic models built from synthesizing sparse amounts of existing flood data are timely, but do not model the underlying hydrodynamics of a flood, limiting simulations to water levels alone. Capturing the full impact of a flood disaster requires models that include strong currents that can harm people, flood momentum that damages infrastructure, and debris flows that can block water outflows and evacuation routes. To address this operational gap, we are developing a near-real-time "nowcast" flood disaster simulation fusing physics-based models with machine learning. As a part of this effort, we are creating training datasets using physics-based hydrodynamic models verified against recent historical flood events in the United States. These data will inform a timely flood model that simulates more than water levels: flow properties such as velocity can improve damage and risk calculations, as well as situational awareness of regions that could be hazardous during a flood. With this nowcast, we aim to build the timely, trustworthy foundation of tools that decision-makers can use to prepare for, respond to, and recover from flood disasters.

  • Open Access

    Equity-Driven Proportional Allocation of Resources (EDPAR) Framework: A Case Study of Georgia Public School Districts and CS AP Teacher Distribution

    by Aaja Christie, Emma Mcdaniel, Chetan Tiwari, Armin R. Mikler, Anu G. Bourgeois

    With the growing demand for computer scientists and information technology personnel, there is increasing recognition of the importance of introducing students to computer science (CS) education throughout their K–12 studies. Across the United States, there is a concerted effort to integrate CS courses into high school curricula and to expose students to CS concepts as early as elementary school. In response, many states have passed legislation mandating that all high schools offer CS courses, under the assumption that increasing the number of course offerings will equate to broader access and opportunity for all students. However, this paper demonstrates that simply expanding the availability of CS courses does not inherently ensure equitable access, particularly when the underlying distribution of qualified teachers and other key resources remains uneven. These policy efforts, while well-intentioned, risk reinforcing or even exacerbating existing disparities in access to quality CS education and, by extension, potential participation in the technical workforce. To address this challenge, we introduce the Equity Driven Proportional Allocation of Resources (EDPAR) framework as both a method for assessing equity in resource distribution and a practical tool for identifying where resources, such as teachers, should be allocated using a student-weighted approach based on current imbalances. EDPAR supports policymakers and educational leaders in making data-informed decisions that go beyond course availability by quantifying systemic inequities in current resource allocation. We present a case study of Georgia public high schools to illustrate the application of EDPAR, focusing specifically on the allocation of Advanced Placement (AP) Computer Science teachers and its relation to the proportion of Historically Marginalized Groups (HMGs) enrolled in each school. Importantly, the EDPAR framework is generalizable and can be applied to guide equitable and proportional allocation of any quantifiable educational resource across schools or districts.

  • Open Access

    Estimating health benefits of a peri-urban railway: A quantitative health impact assessment focused on diabetes

    by Soumya Mazumdar, Nicole Aiken, Aniruddha Banerjee, Andrew Golden, Chetan Tiwari, Fiona Haigh


  • On Self-Improving Token Embeddings

    by Mario M. Kubek, Shiraj Pokharel, Thomas Böhme, Emma L. McDaniel, Herwig Unger, Armin R. Mikler


  • Open Access

    On Self-improving Token Embeddings

    by Mario M. Kubek, Shiraj Pokharel, Thomas Böhme, Emma L. McDaniel, Herwig Unger, Armin R. Mikler

    This article introduces a novel and fast method for refining pre-trained static word or, more generally, token embeddings. By incorporating the embeddings of neighboring tokens in text corpora, it continuously updates the representation of each token, including those without pre-assigned embeddings. This approach effectively addresses the out-of-vocabulary problem, too. Operating independently of large language models and shallow neural networks, it enables versatile applications such as corpus exploration, conceptual search, and word sense disambiguation. The method is designed to enhance token representations within topically homogeneous corpora, where the vocabulary is restricted to a specific domain, resulting in more meaningful embeddings compared to general-purpose pre-trained vectors. As an example, the methodology is applied to explore storm events and their impacts on infrastructure and communities using narratives from a subset of the NOAA Storm Events database. The article also demonstrates how the approach improves the representation of storm-related terms over time, providing valuable insights into the evolving nature of disaster narratives.

  • Open Access

    Zero-Shot Classification of Crisis Tweets Using Instruction-Finetuned Large Language Models

    by Emma McDaniel, Samuel Scheele, Jeff Liu

    Social media posts are frequently identified as a valuable source of open-source intelligence for disaster response, and pre-LLM NLP techniques have been evaluated on datasets of crisis tweets. We assess three commercial large language models (OpenAI GPT-4o, Gemini 1.5-flash-001 and Anthropic Claude-3-5 Sonnet) capabilities in zero-shot classification of short social media posts. In one prompt, the models are asked to perform two classification tasks: 1) identify if the post is informative in a humanitarian context; and 2) rank and provide probabilities for the post in relation to 16 possible humanitarian classes. The posts being classified are from the consolidated crisis tweet dataset, CrisisBench. Results are evaluated using macro, weighted, and binary F1-scores. The informative classification task, generally performed better without extra information, while for the humanitarian label classification providing the event that occurred during which the tweet was mined, resulted in better performance. Further, we found that the models have significantly varying performance by dataset, which raises questions about dataset quality.

  • Open Access

    Unintentional Barriers for AP Computer Science Principles: A Course Designed for Every Student

    by Emma L. McDaniel, Aaja Christie, Anu G. Bourgeois

    In 2016, College Board launched Advanced Placement (AP) Computer Science Principles (CSP) as a college-level introductory course to the breadth of Computer Science (CS). This was a joint effort with National Science Foundation to produce a course with a lower barrier of entry to create new pathways of introducing CS to traditionally underrepresented groups in computing; a high school course designed for every student. The sole recommended prerequisite for the course is Algebra I; however, we found that often additional prerequisites and barriers have been imposed. These additional requirements can inadvertently undermine the inclusive vision that underpinned the course's creation. We focus on public high schools in Georgia and discuss different policy barriers that can impact enrollment in AP CSP.


  • Open Access

    Quantifying the land and population risk of sewage spills overland using a fine-scale, DEM-based GIS model

    by Emma L. McDaniel, Samuel F. Atkinson, Chetan Tiwari

    Accidental releases of untreated sewage into the environment, known as sewage spills, may cause adverse gastrointestinal stress to exposed populations, especially in young, elderly, or immune-compromised individuals. In addition to human pathogens, untreated sewage contains high levels of micropollutants, organic matter, nitrogen, and phosphorus, potentially resulting in aquatic ecosystem impacts such as algal blooms, depleted oxygen, and fish kills in spill-impacted waterways. Our Geographic Information System (GIS) model, Spill Footprint Exposure Risk (SFER) integrates fine-scale elevation data (1/3 arc-second) with flowpath tracing methods to estimate the expected overland pathways of sewage spills and the locations where they are likely to pool. The SFER model can be integrated with secondary measures tailored to the unique needs of decision-makers so they can assess spatially potential exposure risk. To illustrate avenues to assess risk, we developed risk measures for land and population health. The land risk of sewage spills is calculated for subwatershed regions by computing the proportion of the subwatershed’s area that is affected by one modeled footprint. The population health risk is assessed by computing the estimated number of individuals who are within the modeled footprint using fine-scale (90 square meters) population estimates data from LandScan USA. In the results, with a focus on the Atlanta metropolitan region, potential strategies to combine these risk measures with the SFER model are outlined to identify specific areas for intervention.


  • Open Access

    Benchmarking a fast, satisficing vehicle routing algorithm for public health emergency planning and response: “Good Enough for Jazz”

    by Emma L. McDaniel, Sampson Akwafuo, Joshua Urbanovsky, Armin R. Mikler

    Due to situational fluidity and intrinsic uncertainty of emergency response, there needs to be a fast vehicle routing algorithm that meets the constraints of the situation, thus the receiving-staging-storing-distributing (RSSD) algorithm was developed. Benchmarking the quality of this satisficing algorithm is important to understand the consequences of not engaging with the NP-Hard task of vehicle routing problem. This benchmarking will inform whether the RSSD algorithm is producing acceptable and consistent solutions to be used in decision support systems for emergency response planning. We devise metrics in the domain space of emergency planning, response, and medical countermeasure dispensing in order to assess the quality of RSSD solutions. We conduct experiments and perform statistical analyses to assess the quality of the RSSD algorithm’s solutions compared to the best known solutions for selected capacitated vehicle routing problem (CVRP) benchmark instances. The results of these experiments indicate that even though the RSSD algorithm does not engage with finding the optimal route solutions, it behaves in a consistent manner to the best known solutions across a range of instances and attributes.


  • Open Access

    Student Research Abstract: Unsupervised Key Term Extraction of Tornado Narratives from NOAA Storm Events Database | Proceedings of the 38th ACM/SIGAPP Symposium on Applied Computing

    by Emma Louise McDaniel

    Disaster records are often composed of key metrics of the disaster event and its monetary cost. While the cost of the event is helpful for insurance adjustments, the monetary impact of a disaster is not conducive for disaster preparedness planners to build community resilience. Often, there are natural language narratives about the disaster event, like in the National Oceanic and Atmospheric Administration (NOAA) Storm Events Database, that contain essential information but are not easily retrievable because of their unstructured nature. These narratives need to be text mined in order to retrieve the impacts of the disasters in order to structure the data for further use. The method proposed in this abstract is a critical first step in the process. It is an unsupervised key term extraction method using sentence transformers to create embeddings that are then clustered, and assigned key terms by utilizing the highest term frequency-inverse document frequency (tf-idf) scores for the sentences in the narratives.