We introduce PageLocQA, a modular retrieval-augmented generation (RAG) framework tailored for page-level source attribution in long-document question answering. Rather than relying on complex agentic controllers or top-k retrieval alone, PageLocQA adopts a lightweight strategy that combines manual tool routing with structured prompting to generate semantically accurate answers along with page citations.We evaluate PageLocQA on a curated MODTRAN QA dataset, verified by domain experts, as well as benchmark datasets such as HotpotQA, TriviaQA, and Natural Questions. The MODTRAN 6 User Manual is a comprehensive technical manual detailing atmospheric radiative transfer modeling, input parameters, and configuration workflows, making it an ideal testbed for page-grounded QA. While it does not outperform large-scale frameworks like ChatQA or SelfRAG, PageLocQA achieves competitive BERTScore results across all datasets and demonstrates superior grounding performance on MODTRAN, achieving a BERTScore of 0.871 and F1 score of 0.303.These findings underscore the practicality of lightweight RAG-based approaches in settings that require both accuracy and explainability, especially in resource-constrained or scientific QA environments.

Monitoring the indoor mobility of healthcare practitioners is a key factor in improving clinic layout and time management. Traditional methods for tracking practitioners’ indoor mobility rely on subjective self-reports or resource-intensive manual assessments. This paper introduces a novel mobile application designed to automate indoor mobility detection for healthcare practitioners, leveraging smartphone sensors and Bluetooth Received Signal Strength Indicators. The application combines real-time data collection with inter-device data sharing from a swarm of stationary devices to generate detailed mobility profiles. Initial evaluations suggest the application enhances traditional mobility assessments' efficiency and accuracy, offering a scalable and cost-effective solution for healthcare practitioner monitoring. The proposed tool aims to enable clinics with actionable data, improving practitioner efficiency and patient outcomes while reducing overhead associated with manual tracking.

In response to the growing need for socially responsible computer scientists and data scientists, our team is developing a comprehensive data science certificate program specifically tailored for non-computing majors, with a focus on data science for social justice. This program aims to broaden participation in data science and create non-traditional pathways for diverse student populations. Each course in the program is designed to be accessible to non-computing majors, equipping them with the skills to analyze and address social justice issues through data science. Process Oriented Guided Inquiry Learning (POGIL) is employed as an instructional strategy promoting active learning, and real datasets related to social justice are utilized for hands-on activities and assignments, enhancing practical learning experiences. The courses are taught in a synchronous hybrid format, across multiple universities, accommodating both live online and in-person students.

As AI technologies become more prevalent, it is crucial for students to develop responsible, ethical, and proactive AI engagement skills. Recent educational initiatives have focused on enhancing CS and engineering students' AI ethics education but have largely overlooked integrating these concepts across other disciplines. This paper presents and assesses a pioneering initiative that integrates AI ethics into university curricula through a collaborative framework between CS and domain educators. We introduced 1-3 week AI ethics modules in seven diverse courses from Art to Chemistry, incorporating case studies and hands-on activities using chat- or image-based Large Language Models (LLMs). Student surveys indicated significant gains in confidence regarding AI ethics discussions, application of principles, and reasoning skills. Our approach advocates for utilizing structured frameworks and faculty collaboration in embedding AI ethics into university curricula, enhancing students' practical skills and ethical understanding across diverse professional settings.

Wildfires in North America present a grave danger, leading to extensive damage. In this paper, we proposed a deep learning method to rapidly detect wildfires using drones or ground cameras in remote locations. The performance of the proposed system was assessed on public datasets, focusing on accuracy and energy efficiency. The study explores the possibility of running the proposed algorithm on low-power devices like NVIDIA Jetson Nano for real-time fire identification in remote regions.

This BOF welcomes all from Minority Serving Institutions (MSIs) and those with an interest in improving the SIGCSE experience for participants from MSIs. Participants in this BOF will share experiences and collectively identify ways to effect greater inclusion in SIGCSE TS of those from Minority Serving Institutions and to foster growth in future participation. In addition, this BOF session will serve to connect participants from and interested in MSIs.

Generating paths of a mobile device in indoor space by sensing its Bluetooth RSSI value is challenging but has real-world applications. Although Bluetooth RSSI suffers from different factors that limit its usability, this research shows that it can still be used to detect mobility and, over a duration of time, can be used to form paths. This poster presents algorithms that can create a path of a moving mobile device by sensing its RSSI values over time and then presents early results of the algorithm's effectiveness while tracking health practitioners' movement within a community care clinic setting.

Students' conceptual knowledge about parallelism continues to grow because of the increasing infusion of parallelism topics in all courses, even if they are not dedicated to those topics. Most times, such infusion is through high-level theoretical problems, as listed in the textbooks. However, without providing corresponding real-life instances, such problems might not engage students with the content in such courses. Does providing real-life activities along with theoretical constructs help students to grasp such topics better? As an attempt to explore student comprehension of parallelism in such course scenarios, this paper describes an unplugged assignment, where students were asked to categorize their daily activities into concurrent and parallel tasks. The paper then presents a thematic analysis of student experience using three semesters' worth of student data. The analysis shows that, although students exhibit limited knowledge of parallelism in real-life situations and have numerous misconceptions about parallelism; their participation in such assignments are deliberate, their thinking processes are analytical, their writing skills in expressing such concept is limited and their impressions about such assignments are positive. The paper then presents an effort to enhance student's comprehension and shows how the intervention has improved students' understanding of this critical concept. This study confirms that it is essential to provide real-life analogies and examples to facilitate a greater understanding of parallelism. Because providing only theoretical examples might not engage a diverse body of students in courses that are just trying to expose students to this critical topic.

Big data and cloud computing collectively offer a paradigm shift in the way businesses are now acquiring, using, and managing information technology. This creates the need for every CS student to be equipped with foundational knowledge in this collective paradigm and possess some hands-on experience in deploying and managing big data applications in the cloud. This study argues that, for substantial coverage of big data and cloud computing concepts and skills, the relevant topics need to be integrated into multiple core courses across the CS curriculum rather than creating additional courses and performing a major overhaul of the curriculum. Our approach to including these topics is to develop autonomous competency-based learning modules for specific core courses in which their coverage might find an appropriate context. In this paper, four such modules are discussed, and our classroom experiences during these interventions are documented. Student performance data and survey results show reasonable success in attaining student learning outcomes, enhanced engagement, and interests.

With the growing popularity of smartphones, continuous and implicit authentication of such devices via behavioral biometrics such as touch dynamics becomes an attractive option, especially when the physical biometrics are challenging to utilize, or their frequent and continuous usage annoys the user. However, touch dynamics is vulnerable to potential security attacks such as shoulder surfing, camera attack, and smudge attack. As a result, it is challenging to rule out genuine imposters while only relying on models that learn from real touchstrokes. In this paper, a touchstroke authentication model based on Auxiliary Classifier Generative Adversarial Network (AC-GAN) is presented. Given a small subset of a legitimate user's touchstrokes data during training, the presented AC-GAN model learns to generate a vast amount of synthetic touchstrokes that closely approximate the real touchstrokes, simulating imposter behavior, and then uses both generated and real touchstrokes in discriminating real user from the imposters. The presented network is trained on the Touchanalytics dataset and the discriminability is evaluated with popular performance metrics and loss functions. The evaluation results suggest that it is possible to achieve comparable authentication accuracies with Equal Error Rate ranging from 2% to 11% even when the generative model is challenged with a vast number of synthetic data that effectively simulates an imposter behavior. The use of AC-GAN also diversifies generated samples and stabilizes training.

In light of recent local, national and global events, spatial justice provides a potentially powerful lens by which to explore a multitude of spatial inequalities. For more than two decades, scholars have been espousing the power of spatial justice to help develop more equitable and just communities. However, defining spatial justice and developing a methodology for quantitatively analyzing spatial justice is complicated and no agreed upon metric for examining spatial justice has been developed. Instead, individual measures of spatial injustices have been studied. One such individual measure of spatial justice is economic mobility. Recent research on economic mobility has revealed the importance of local geography on upward mobility and may serve as an important keystone in developing a metric for multiple place-based issues of spatial injustice. As a result, this paper seeks to explore place-based variables within individual census tracts in an effort to understand their impact on economic mobility and potentially spatial justice. The methodology relies on data science and machine learning techniques and the results show that the deep leaning model is able to predict economic mobility of a census tract based on its spatial variables with 89% accuracy. In the end, this research will allow for comparative analysis between differing geographies and also identify leading variables in the overall quest for spatial justice.

Data Science is an essential concept for twenty-first century workforce and as a result the need to help all students acquire such skill has recently gained increased attention. However, most smaller schools are currently facing challenges to provide related knowledge and skill to a broad student population. During Spring of 2019, we organized a faculty workshop on "Data Science Pedagogy and Practice" aimed at building and enhancing data science capacity (teaching, research, partnership, collaboration) at our institution. We gathered faculty input on what would be needed for a successful university-wide initiative to incorporate data analytics concepts as a basic component of training across variety of disciplines including science, business and social sciences. We were able to identify certain challenges and opportunities to accommodate deeper coverage of data science in the undergraduate teaching and research, and as a result, we initiated few efforts such as 1) teaching through faculty partnership, 2) module-based integration into existing courses rather than developing new courses, 3) development of reusable course modules and augment that with contextual hand-on projects so that students could appreciate the use of data science in their own career path, 4) facilitation of course preparation and implementation via a small faculty adopter grant, and 5) development of a graduate certificate program in data analytics. In this poster, we report our experience in organizing and implementing the workshop, the key aspects of the university-wide data science efforts initiated as a result of the workshop, and the lessons learned so far from these initiatives.

In this poster, we describe our effort to develop, pilot, and evaluate a model for infusing data literacy into undergraduate curricula across a variety of disciplines using a modular approach. Our pilot implementation achieved reasonable success in attaining student learning outcomes, enhanced engagement, and interests.

Big data and cloud computing collectively offer a paradigm shift in the way businesses are now acquiring, using and managing information technology. This creates the need for every CS student to be equipped with foundational knowledge in this collective paradigm and to possess some hands-on experience in deploying and managing big data applications in the cloud. We argue that, for substantial coverage of big data and cloud computing concepts and skills, the relevant topics need to be integrated into multiple core courses across the undergraduate CS curriculum rather than creating additional standalone core or elective courses and performing a major overhaul of the curriculum. Our approach to including these topics is to develop autonomous learning modules for specific core courses in which their coverage might find an appropriate context. In this paper, three such modules are discussed and our classroom experiences during these interventions are documented. So far, we have achieved reasonable success in attaining student learning outcomes, enhanced engagement, and interests. Our objective is to share our experience with the academics who aim at incorporating similar pedagogy and to receive feedback about our approach.

Classroom formative assessment augmented with timely and frequent feedback has become one of the most prominent teaching practices in education research. On the context of Computer Science (CS) courses that expose students to the functionality and dynamic aspects of various algorithms, traditionally, students are evaluated by exploring in-class paper-based exercises. In these exercises, they simulate the steps of an algorithm by drawing several instances of a diagram. This traditional approach is time consuming, is inherently difficult for students to express the dynamics of an algorithm, does not allow timely feedback, and restricts the number of exercises that students can practice and receive feedback on. Mobile Response System (MRS) is a software environment that facilitates in-class exercises and their real-time assessment using mobile devices and therefore focuses on addressing many of the above-mentioned problems. In this paper, we present results of eight semester-long studies using MRS in two of the required CS courses at Winston-Salem State University (WSSU). Our experimental evaluation shows the educational benefits of the proposed approach in terms of enhanced student retention of covered concepts, reduced failing rate, and increased student engagement and satisfaction.

No abstract available.

Big data and cloud computing (BDCloud) collectively offer a paradigm shift in the way businesses are now acquiring, using and managing information technology. With the fast growth of this paradigm, we argue that each and every CS and IT students should be equipped with foundation knowledge in this collective paradigm and should possess hand-on-experiences in managing big data applications in clouds to acquire skills that are necessary to meet current and future industry demands. This poster presents our research that proposes gradual and systematic integration of big data and cloud computing related topics into multiple core (required) courses of CS/IT curriculum. The poster, supported by a NSF grant, will be useful for CS/IT students and their instructors as it identifies big data and cloud computing related topics that are important to cover, finds a sequence of the prescribed topics that can be incorporated into existing core courses most effectively, and suggests specific core courses in which their coverage might find an appropriate context. The poster further identifies the major challenges this proposed intervention may encounter and provides a deeper analysis of them. Finally, the poster describes our experience of implementing one such course with proposed interventions during Fall of 2016 semester. The pre- post- test results that measure student opinion and understanding of big data and cloud computing topics are presented in the poster and demonstrate improved student interest and learning.

Computer Science instructors have been exploiting learning technology such as Algorithm Visualization (AV) for last few years to explain hard-to-understand algorithms to the learners through simulations and animations. In this work, we explore an active and highly engaging approach, namely, the construction of visualizations of the algorithms under study. Our approach is further augmented with automated assessment of students' in-class construction activities, which they execute as apps in their mobile devices. In this paper, we utilize case study, a step-by-step visualization of a construction exercise app, to explain how technology is leveraged to provide a richer way for learners to interact with a problem, and how instructor can acquire real-time evidence of learners' comprehension of covered lecture material. Our experimental evaluation shows the educational benefits of the proposed approach in terms of enhanced student learning, reduced drop-out rate and increased student satisfaction.

Over the past couple of years, evidence-based teaching and learning methods are brought into focus from the experience gained in clinical psychology and their use of Evidence-Based Practices. Different authors have discussed the advantages of using such evidence-based methods for teaching and learning in academia. Measuring real-time impact of traditional pedagogical approaches used in STEM disciplines are not easy and do not provide faculty an instant evidence about student learning. This paper will present Mobile Response System (MRS) software, which facilitate anonymous communication, interaction and evaluation of in-class interactive problem solving activities using mobile devices. MRS facilitates a feedback-driven and evidence-based teaching methodology, which is important to enhance student learning.

To improve student's class experience, the use of mobile devices has been steadily increasing. However, such use of mobile learning environments in the class is mostly static in nature through content delivery or multiple choice and true/false quiz taking. In CS courses, we need learning environments where students can interact with the problem in a hands-on-approach and instructor can assess their learning skills in real-time with problems having different degree of difficulty. To facilitate such interactive problem solving and real-time assessment using mobile devices, a comprehensive backend system is necessary. This paper presents one such system, named Mobile Response System (MRS) software, associated interactive problem-solving activities, and lessons learned by using it in the CS classrooms. MRS provides instructor with the opportunity of evidence-based teaching by allowing students to perform interactive exercises in their mobile devices with different learning outcomes and by getting an instant feedback on their performance and mental models. MRS is easy-to-use, extensible and can render interactive exercises developed by third-party developers. The student performance data shows its effectiveness in increasing student understanding of difficult concepts and the overall perception of using the software was very positive.

A recent Pew research center study of mobile device usage revealed that, African American and Latinos are the most active users of the Internet from mobile devices. The study also revealed that minority cell phone owners take advantage of a much greater range of their phone's features compared with people of other ethnicities. At Winston Salem State University (WSSU), it is common for students to multitask and use their mobile devices while in class for studying, or performing other activities. This paper reports our ongoing experiences running a National Science Foundation (NSF)-sponsored targeted Infusion Project (TIP) in Computer Science Department that aims to leverage this situation by developing a mobile classroom response system (MRS) to allow students solve interactive problems in their mobile devices in order to improve their class engagement and problem solving skills. By allowing them to solve problems in their preferred devices, the project expects to create a friendly learning environment where the students want to retain, be active and skillful.

Evidence-based instructional practices were incorporated in class, which gave immediate indication on student's problem solving skills and class participation information. This pedagogy showed positive results and broader acceptance by students in several semesters of intervention. Significant usage of mobile devices during class motivates the extension of this pedagogical approach of asynchronous problem solving using mobile devices. We believe that use of such devices in the classroom for solving interactive problems will enhance student's abilities to solve problems by using their preferred interaction mode. This paper presents the results of the evidence based pedagogy and development of a mobile classroom response system that extends this pedagogy to help student solve interactive problems in their mobile devices to improve their class engagement and problem solving skills.

This poster addresses a significant learning barrier experienced at many CS departments, specially at predominantly minority institutions, which is the problem of students? inability to keep engaged and interested in classroom. In this research, we investigate the applicability of using mobile devices in the classroom and incorporation of interactive problem solving using those devices to increase class engagement and active learning for students. By allowing the students to solve problems in their preferred devices, the research expects to create a friendly learning environment where the students want to retain, be active and skillful. The poster will present the design aspects of Mobile Response System (MRS) software that will be utilized to communicate, collaborate and evaluate interactive problems using mobile devices. The poster will also showcase several interactive problem-solving activities utilizing mobile devices and MRS software, which have been developed and are being adopted in CS and IT courses at Winston-Salem State University (WSSU). It is expected that this research will invigorate interest in Computer Science among minority and underrepresented students through exposure to the technology-rich learning environment. By enhancing student learning and problem solving abilities, it is also expected that this research work will improve the quality and quantity of underrepresented minority students in STEM workforce or graduate study. The successful execution of this project will advance research and the knowledge of mobile device usage in CS classrooms and more importantly the way it impact teaching strategy and student learning at WSSU and other institutions.

Recent research indicates that women and minority students find computer science more meaningful and engaging when they have a chance to apply their knowledge within communal and societal context. We report the experience gained while incorporating a software engineering project that aid community and society at a university that predominantly serves underrepresented minority students. The project that we utilized allows the students to apply the theories and principles of software engineering on a real life scenario while keeping them engaged and motivated, addresses the needs of community and society, and better prepares the minority students for their professional career through improved academic achievement, enhanced self-reliance and community engagement. The project can easily be replicated and adopted to any project-based software engineering course taught at any university and likely to generate similar benefits to students and society that we noticed.

Document clustering has become an increasingly important task in analyzing huge numbers of documents distributed among various sites. The challenging aspect is to analyze this enormous number of extremely high dimensional distributed documents and to organize them in such a way that results in better search and knowledge extraction without introducing much extra cost and complexity. This paper presents a distributed document clustering approach called distributed information bottleneck (DIB). DIB adopts a two stage agglomerative information bottleneck (aIB) algorithm to generate local clusters. At the first stage, the high-dimensional document vector is significantly reduced by finding word-clusters. These word-clusters are then used to obtain document-clusters in the second stage. DIB then extracts compact but informative local models from these document-clusters and transfers them to a central site. At the global site, the local models, that are likely to describe the same document set, are first combined. The resultant local models are then clustered by using the aIB algorithm to produce a hierarchical organization of all distributed documents. Our experimental results demonstrate the robustness, efficiency and effectiveness of DIB approach to cluster distributed documents.