Theses and Dissertations from UMD
Permanent URI for this communityhttp://hdl.handle.net/1903/2
New submissions to the thesis/dissertation collections are added automatically as they are received from the Graduate School. Currently, the Graduate School deposits all theses and dissertations from a given semester after the official graduation date. This means that there may be up to a 4 month delay in the appearance of a give thesis/dissertation in DRUM
More information is available at Theses and Dissertations at University of Maryland Libraries.
Browse
15 results
Search Results
Item COMPLEXITY CONTROLLED NATURAL LANGUAGE GENERATION(2023) Agrawal, Sweta; Carpuat, Marine; Computer Science; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Generating text at the right level of complexity for its target audience so it can be easily understood by its target audience has the potential to make information more accessible to a wider range of people, including non-native speakers, language learners, and people who suffer from language or cognitive impairments. For example, a native Hindi speaker learning English might prefer reading a U.S. news article in English or Hindi tailored to their vocabulary and language proficiency level. Natural Language Generation (NLG), the use of computational models to generate human-like text, has been used to empower countless applications – from automatically summarizing financial and weather reports to enabling communication between multilingual communities through automatic translation. Although NLG has met some level of success, current models ignore that there are many valid ways of conveying the same information in a text and that selecting the appropriate variation requires knowing who the text is written for and its intended purpose. To address this, in this thesis, we present tasks, datasets, models, and algorithms that are designed to let users specify how simple or complex the generated text should be in a given language. We introduce the Complexity Controlled Machine Translation task, where the goal is to translate text from one language to another at a specific complexity level defined by the U.S. reading grade level. While standard machine translation (MT) tools generate a single output for each input, the models we design for this task produce translation at various complexity levels to suit the needs of different users. In order to build such models, we ideally require rich annotation and resources for supervised training, i.e., examples of the same input text paired with several translations in the output language, which is not available in most datasets used in MT. Hence, we have also contributed datasets that can enable the generation and evaluation of complexity-controlled translations. Furthermore, recognizing that when humans simplify a complex text in a given language, they often revise parts of the complex text according to the intended audience, we present strategies to adopt general-purpose Edit-based Non-Autoregressive models for controllable text simplification (TS). In this framing, the simplified output for a desired grade level is generated through a sequence of edit operations like deletions and insertions applied to the complex input sequence. As the model needs to learn to perform a wide range of edit operations for different target grade levels, we introduce algorithms to inject additional guidance during training and inference, which results in improved output quality while also providing users with the specific changes made to the input text. Finally, we present approaches to adapt general-purpose controllable TS models that leverage unsupervised pre-training and low-level control tokens describing the nature of TS edit operations as side constraints for grade-specific TS. Having developed models that can enable complexity-controlled text generation, in the final part of the thesis, we introduce a reading comprehension-based human evaluation framework that is designed to assess the correctness of texts generated by these systems using multiple-choice question-answering. Furthermore, we evaluate whether the measure of correctness (via the ability of native speakers to answer the questions correctly using the simplified texts) is captured by existing automatic metrics that measure text complexity or meaning preservation.Item Information Olfactation: Theory, Design, and Evaluation(2019) Patnaik, Biswaksen; Elmqvist, Niklas; Computer Science; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Olfactory feedback for analytical tasks is a virtually unexplored area in spite of the advantages it offers for information recall, feature identification, and location detection. Here we introduce the concept of ‘Information Olfactation’ as the fragrant sibling of information visualization, and discuss how scent can be used to convey data. Building on a review of the human olfactory system and mirroring common visualization practice, we propose olfactory marks, the substrate in which they exist, and their olfactory channels that are available to designers. To exemplify this idea, we present ‘viScent(1.0)’: a six-scent stereo olfactory display capable of conveying olfactory glyphs of varying temperature and direction, as well as a corresponding software system that integrates the display with a traditional visualization display. We also conduct a comprehensive perceptual experiment on Information Olfactation: the use of olfactory marks and channels to convey data. More specifically, following the example from graphical perception studies, we design an experiment that studies the perceptual accuracy of four ``olfactory channels''---scent type, scent intensity, airflow, and temperature---for conveying three different types of data---nominal, ordinal, and quantitative. We also present details of an advanced 24-scent olfactory display: ‘viScent(2.0)’ and its software framework that we designed in order to run this experiment. Our results yield a ranking of olfactory channels for each data type that follows similar principles as rankings for visual channels, such as those derived by Mackinlay, Cleveland & McGill, and Bertin.Item Menthol Tobacco Sales Restrictions at the Local Level: Community Perceptions, Lessons Learned, and Policy Evaluation(2019) D'Silva, Joanne; Fryer, Craig S.; Public and Community Health; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Menthol cigarettes are a significant public health concern. While sales of regular cigarettes have declined in recent years, menthol cigarette sales are increasing. Known for their minty taste and cooling effects, menthol cigarettes are a starter product for youth. The tobacco industry has a long history of predatory marketing to African Americans, who are more likely to smoke menthol cigarettes, have lower quit rates, and experience higher tobacco-related mortality than white smokers. Because flavored cigarettes are appealing to youth, the 2009 Tobacco Control Act banned the sale of flavored cigarettes; however, menthol was exempted. Due to federal inaction on menthol, localities are enacting restrictions to address this social justice issue. In 2017, Minneapolis, Minnesota was among the first U.S. jurisdictions to restrict menthol sales to tobacco shops and liquor stores. Grounded in the Social Ecological Model, this dissertation utilized multiple methods to increase our understanding of the experience and impact of the Minneapolis menthol policy. Study 1 examined African American adult smoker perceptions and found that many menthol smokers perceived menthol to be as harmful or more harmful than non-menthol cigarettes and emphasized the role of tobacco industry targeting to African American communities and youth. Participants also indicated mixed support for menthol policy restrictions. Study 2 identified key factors that led to policy passage and included an in-depth assessment of critical steps in policy-making from policy formulation through adoption. The active engagement of youth and support of members of communities most impacted by menthol tobacco-related disparities were identified as critical to the policy’s successful adoption. Study 3 evaluated the impact of the menthol policy on the retail environment and found reduced menthol availability and decreased menthol marketing at the point-of-sale two months following policy implementation. Unintended consequences of the policy were also observed; two tobacco shops were added to convenience stores and continued to sell menthol tobacco. Results from these studies provide critical insights from several phases of the policy process. Findings can inform other jurisdictions around the country interested in pursuing menthol restrictions to reduce the burden of tobacco use and advance health equity for priority populations.Item Evaluation of an Alternative School's Impact on the Graduation Rate Overall and for Students Receiving Free or Reduced Price Meals in One Local School System(2017) Zimmerman, Wendy Lynn; Imig, David; Education Policy, and Leadership; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Free and reduced meals (FARM) students in one Maryland school system are dropping out of school at a rate almost five times greater than non-FARM students. In order to address the overall and the FARM subgroup dropout rate an intervention program was implemented. The program invited students with the greatest risk of dropping out to attend. Small class sizes and faculty that focused on building relationships and meeting each individual student’s social and emotional needs are hallmarks of the intervention. The effectiveness of the program was established through three tests: 1. Finding the average overall dropout rate before the inception of the program compared to dropout rate after the implementation; 2. A logistic regression to determine the probability of a student graduating from high school based on data from a group of students who attended the intervention as compared to a demographically matched group of students who did not attend; 3. A logistic regression to determine the probability of a student graduating based on data from a group of students who attended the program as compared to a pooled group of students who where invited to attend, but did not. The results suggest the program is effective; however, the reader should be cautioned as the results are based on a small sample size. The county experienced a 5.35% decrease in the overall dropout rate and a 6.80% decrease in the FARM dropout rate after the implementation of the program. Matched students who attended the program were almost fifteen times more likely to graduate than their peers who did not attend the program, and matched FARM students were fourteen times more likely to graduate. Students who attended the program had an almost nine times greater chance of graduating, and FARM students had an eight times greater probability of graduating, than students who were invited to attend but did not. The results show a relationship between the implementation of the dropout intervention program and a decrease in the dropout rate for the county and for the FARM subgroup.Item Development of Planning And Evaluation Models For Superstreets(2016) Xu, Liu; Chang, Gang-Len; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Despite the extensive implementation of Superstreets on congested arterials, reliable methodologies for such designs remain unavailable. The purpose of this research is to fill the information gap by offering reliable tools to assist traffic professionals in the design of Superstreets with and without signal control. The entire tool developed in this thesis consists of three models. The first model is used to determine the minimum U-turn offset length for an Un-signalized Superstreet, given the arterial headway distribution of the traffic flows and the distribution of critical gaps among drivers. The second model is designed to estimate the queue size and its variation on each critical link in a signalized Superstreet, based on the given signal plan and the range of observed volumes. Recognizing that the operational performance of a Superstreet cannot be achieved without an effective signal plan, the third model is developed to produce a signal optimization method that can generate progression offsets for heavy arterial flows moving into and out of such an intersection design.Item PERFORMANCE AND QUALITY INDICATOR DATA AVAILABILITY FOR P3 PROJECTS(2015) Robinson, Matthew Christian; Cui, Qingbin; Civil Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)Public Private Partnerships (P3’s) are a relatively new project delivery method. In reaction to the increased attention to this delivery method, the Federal Highway Administration has begun to enhance its existing P3-VALUE toolkit, an educational toolkit which demonstrates the potential benefits of utilizing a Public Private Partnership over a traditional method like Design Bid Build for transportation projects. The toolkit utilizes assumptions about P3 project characteristics to build the scenarios for its analyses. Unfortunately, there is a significant lack of data which would serve to justify assumptions made about improved P3 quality performances. Furthermore, there is a basic lack of knowledge regarding what data is even available to make certain assumptions. The intent of this thesis is to identify what data can be collected, what data can be shared, and to determine what data can be expected to be reliably available, and not subject to proprietary rights, for future analysis regarding the improved P3 quality performance.Item An Analysis of the Academic Growth of Students with Disabilities Instructed in Self-Contained Schools and a Unique Therapeutic Inclusive Public School Special Education Program(2014) Gladhill, Jeffrey Scott; Kivlighan, Dennis M; Education Policy, and Leadership; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This study analyzes the academic growth of students with disabilities (SWD) educated in two self-contained schools and a unique therapeutic inclusive public school program (TP) located in Piedmont County (PC). The majority of participants in this study were classified as students with an emotional disability (ED) (40%) or other health impairment (OHI) (34%). Results revealed that students in the TP group showed overall higher significant growth in mathematics as gauged by the Measures of Academic Progress (MAP) assessment, until variables of grade level, gender, and medical assistance (MA) eligibility were added to the analysis. Following an additional analysis of growth accounting for these additional variables, there were no significant differences between the two educational environments. This study suggests that variables of teacher efficacy, relationship, background, prior knowledge, professional development, lack of random assignment of students, teacher certification and content knowledge of teachers need to be studied in future research. Findings from this research should not be generalized beyond this study due to a low number of participants (48 students) and the unique nature of the TP. Additional limitations regarding generalizability and recommendations are addressed within this study.Item Evaluation-Focused Reliability Test Program Planning Methodology(2013) Tamburello, Robert Nicholas; Herrmann, Jeffrey W; Mechanical Engineering; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)In practice, various ad hoc approaches for designing reliability test programs have been observed. Many of these approaches rely on previously established rules of thumb for which the underlying rationale is indefensible. As a consequence, those who use such approaches are unlikely to maintain a firm resource commitment for the conduct of reliability test program activities. Furthermore, it is difficult to ascertain the impact that budgetary cuts will have on the adequacy of the reliability test program with any degree of accuracy. The contributions of this research are as follows. This dissertation presents a novel 7-step planning process to aid practitioners in designing adequate reliability test programs. This planning process serves as a tool to systematically identify, quantify, and mitigate evaluation risks subject to resource constraints. By performing the 7 steps associated with this planning process, practitioners will be able to logically justify reliability test program requirements and more effectively articulate the significance of evaluation risks associated with a particular reliability test program design. Additionally, it is a straightforward process to assess the impact of a reduction in reliability test program resources. This planning process includes a step for assessing the level of risk associated with key aspects of the reliability test program. One such consideration that is of paramount importance is the adequacy of the test configuration of the system. Hence, we present a simulation-based approach for assessing the adequacy of the test configuration of a complex system-of-systems. For the purpose of demonstration, an application of this approach to air defense systems is included; however, the approach is valid for any type of system. As well, this dissertation presents an evaluation risk assessment process for reliability test programs--adapted from the traditional failure mode and effects analysis (FMEA) process. This process can be applied to any reliability test program, irrespective of the manner in which the plan was formulated. Just as a FMEA facilitates the identification of potential weaknesses in a system architecture, this evaluation risk assessment process is designed to surface reliability test program weaknesses and gauge the potential impact of each weakness to the system reliability evaluation.Item VIRTUAL YOUTH SPACES IN PUBLIC LIBRARIES: DEVELOPING AN EVALUATIVE FRAMEWORK(2013) Valdivia, Claire Suzanne; Subramaniam, Mega; Library & Information Services; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)This study explores the successes and challenges of the virtual space component of the 21st Century Learning Labs in Libraries and Museums funded by the Institute for Museum and Library Services (IMLS) and the MacArthur Foundation. This study synthesizes the literature on informal learning for youth that motivated these learning labs and analyzes existing evaluative approaches to public library initiatives. The author adopts a grounded theory approach and conducts semi-structured interviews with learning lab staff from four sites that received the 21st Century Learning Labs in Libraries and Museums Grant. Key concepts from the literature guide the analysis of the interviews to produce a foundational evaluative framework that can be used by public library staff to design and evaluate goals for their virtual learning lab. This framework considers both the overall mission of virtual learning labs and contextual factors that influence the library.Item CROSS-CULTURAL COMPETENCE TRAINING EFFECTIVENESS: THE IMACT OF TRAINING TRANSFER AND PREDICTORS OF TRANSFER(2013) Feinberg, Emily Greene; Hanges, Paul; Psychology; Digital Repository at the University of Maryland; University of Maryland (College Park, Md.)In order to better understand the many inconsistencies found in previous research on cross-cultural competence (3C) training effectiveness, this study investigated the impact of training transfer (i.e., the extent to which training concepts are applied to the job context) on cross-cultural success outcomes. Predictors of training transfer, unique to 3C training setting, were also examined. Specifically, individual characteristics, such as personality (i.e., openness to experience, tolerance for ambiguity), cultural motivation, and previous international experience, in addition to climate for transfer (home and host organizations), were discussed as important factors that will likely impact the extent to which training transfer occurs. Hypotheses were tested in a sample of Peace Corps volunteers (N=101) who were completing a cross-cultural training course in preparation for foreign assignment. Results found support for training transfer main effects, as well as cultural motivation and home organization climate for transfer as significant predictors of transfer. Theoretical and practical implications are discussed.