Teaching computing at scale necessitates some level of automation. Many undergraduate courses have scaled up through commercial Learning Management Systems or, sometimes, ''homegrown'' tools developed by teaching staff. The development of homegrown tools within universities, especially those that lie in computing classrooms, has been reported on for some time. However, research focuses strongly on the applications of these tools, rather than the design process and the overarching ecosystem of software development in which these tools are embedded. We investigate the design of homegrown tools within a large R1 institution, where much of the development is conducted by and for Teaching Assistants (TAs). Our results are synthesized into a four phase design cycle–a framework that allows us to identify gaps in development and grounds future exploration of tool development within computer science education.

Large, high-enrollment computing courses are adopting more structured approaches to supporting students, but questions remain about which structures effectively support teaching and learning. We present a mixed-methods investigation of a large CS2 course that leverages a Student Support System (SSS), which combines flexible extensions with one-on-one support meetings. Through behavioral profiling of student engagement and thematic analysis of support staff interviews, we find that while the SSS expanded access, fostered human connection, and reached underrepresented students, it failed to produce equitable academic outcomes. Our findings reveal the limits of reactive, human-dependent support at scale and highlight suggestions for redesigns. We invite further discussion on where well-intentioned systems like the SSS can fall short and what improvements are needed.

The expansion of undergraduate CS programs brings different forms of student identity, sociotechnical perspectives, and intersectionality into the classroom. These background factors affect student understanding of the world, and, consequently, their work in computing ethic classes. Instructors of computing ethics courses therefore must facilitate topics that are not only pertinent to modern technologies but that are also interesting for students from a range of backgrounds. In this work, we introduce a low-overhead, natural language processing tool that can assist instructors in extracting student talking points from over 600 discussion forum posts in a large-scale undergraduate computing ethics course. When compared to large language model approaches, this n-gram-based scripting tool is more effective in selecting popular quotes and summarizing course discussion. This tool is simple in implementation and can be easily adapted by instructors to prepare for classroom discussion.

Flexible, competency-based learning in large, high-workload undergraduate CS courses provide students options to leverage time, instructors, and other course resources for individualized learning. However, at scale the wide range of resources can overwhelm students and instead create impersonal classrooms. We describe our experiences with a structured student support system that helps students navigate learning in a large CS2 course. This study extends prior work on automated, flexible assignment extensions and introduces a structured approach to scheduling student support meetings, where staff suggest specific assignment work plans and course resources with individual students. Our work-in-progress findings suggest that student support meetings create otherwise missing channels of communications within the large class. We introduce a comprehensive profiling approach based on student support system usage to characterize the background and work patterns of students and how they leverage larger framework of course resources available. This work invites further discussion to understand the complex interactions between undergraduate student learning and adequate course support.

This work explores computing ethics education through a sociological lens, focusing on education's dual role in reflecting and challenging Silicon Valley's hegemonic force. We present a case study of a one-unit computing ethics course at a R1 public university. Discussion-based assignments can foster accessible ethical discussions and scaffold "water cooler" talk among students; these informal conversations provide a starting point for critical engagement with ethical dilemmas. However, as a standalone offering, the course can reinforce the perception of ethics as secondary to technical skills, further highlighting the need to reimagine an embedded computing ethics education that better prepares students to critically engage with and reshape sociotechnical systems.

Many undergraduate computing programs employ undergraduate teaching assistants (UTAs), who are uniquely positioned to positively contribute to classroom and student retention. In this exploratory study, we study UTAs who enrolled in a data science professional pedagogy training course. In the first semester, we collected UTA survey responses and journal entries, which were assignments in the pedagogy course; in the second semester, we interviewed UTAs and observed their classrooms. We examine how awareness of positionality - a critical theme of the pedagogy course - informed teaching goals and practices. We perform a qualitative analysis to explore the following research questions: RQ1: How does UTA identity and/or academic and personal experiences influence their teaching goals? RQ2: How do the teaching goals in RQ1 translate to teaching practices?

This study examines the challenges and opportunities of teaching computing ethics within the context of a large, low-workload, standalone course. CS199 is a one-unit, pass/fail computing ethics course designed to provide students with exposure to a wide array of topics and promote critical peer-based engagement. We leverage submitted work via Question, Quote, Comment, and Replies (QQCRs) and podcasts to facilitate discussions outside the classroom. While QQCRs have shown promise in promoting engagement and exposing students to diverse perspectives, limitations remain in stimulating deeper critiques of the material. We reflect on the effectiveness of asynchronous discussion and its alignment with broader learning goals in computing ethics education.

This study investigates how large language models (LLMs) can identify the presence of learning outcomes within student submitted work in a computing ethics course. To do so, we craft a codebook to spot key learning outcomes, such as the usage of critical reasoning and awareness of various social issues. We leverage the GPT-4o and GPT-3.5-turbo LLMs to apply codes onto 8,500 pieces of student submitted work. We then use Cohen's kappa to assess interrater reliability and compare human reviewers' coding to outputs from those models, finding that GPT-4o performed just as well as the agreement between human reviewers. We then use the model outputs to identify specific course readings that students engaged particularly deeply with to better inform our computing ethics instruction.

Teaching assistants (TAs) are well-poised to improve student performance and retention-particularly for those from minoritized backgrounds. In this experience report, we present a TA pedagogy course that trains TAs to effectively teach large-scale undergraduate courses in data science-and does so by centering inclusion and the classroom as a social space of engagement. This curriculum is grounded in a justice-oriented pedagogy framework and also tackles the interdisciplinary teaching skills necessary to teach data science. Course topics cover pedagogical frameworks and evidence-based teaching practices alongside professional and management skills that reflect the TA's positionality as both instructor and student. We present our experiences with offering this semester-long course and discuss how to adapt this course to other higher education institutions.

This Work-In-Progress Innovative Practice paper describes a new flexible, at-scale assignment extension policy and its implementation in large undergraduate computing class-rooms. While prior work has studied flexible deadlines and their effect on student learning, such policies are still under-utilized in post-secondary classrooms due to practical constraints on administrative workload, from managing hundreds to thousands of requests to reducing excessive grading overhead, especially in large post-secondary classrooms. The Flextensions tool-an automated flexible extension assignment software-promotes equitable learning opportunities in higher education by providing sufficient accommodations to each student's unique learning needs and life circumstances. Part of Flextensions is a scalable software tool implementation that facilitates instructor management of extension requests across thousands of students and a variety of course policies. We present Flextensions and describe initial experiences with adapting the tool to computer science and data science undergraduate courses at an R1 institution in the United States. This work shares the open-source software that enables and streamlines the management of extension requests across different course policies. Additionally, we provide an initial analysis of flexible extensions in three specific large-scale (500–1500 students) undergraduate computing courses. Overall, students tended not to take advantage of the policy but rather used it only when needed, with many citing extenuating circumstances-personal or otherwise. By analzing survey results, the policy was well-received, with positive impressions on well-being, learning outcomes, and overall academic experience. One student-reported benefit was that many felt valued as individuals in the classroom. Despite some students still reporting stigma towards requesting an extension, Flextensions has the promising ability to improve the quality and responsiveness of creating accommodations in large-scale classrooms.

This Work-In-Progress Research paper examines the measurable impacts of a flexible extension policy on course learning objectives in large-scale computing courses. In higher education, flexible extension policies have become increasingly common, where students can individually request additional time on assignments to accommodate their unique learning needs and life circumstances. This paper analyzes the effects of extensions on cultivating student learning and academic performance using flexible extension data from an undergraduate Data Science course at a U.S. Rl institution. We study how extension policies impact different groups of students based on their usage. Using the policy, students are achieving high rates of assignment submissions. While prior experience has no bearing on how students use the extension policy, students who use the policy tend to have slightly lower final exam scores. This early work aims to inform educators about the efficacy of flexible extensions and how they impact student learning and academic outcomes. Ultimately, our goal is to contribute to the creation of a supportive learning environment where all students can succeed. In this paper, we intend to answer the following research questions: 1) How does a flexible extension policy improve students' learning experience in the course? 2) How does student extension usage across a course term correlate with academic learning goals?

Today's rapidly evolving educational landscape prioritizes active student engagement. Classrooms at scale face particular challenges in fostering meaningful interactions between students and course content. In this study, we introduce EIT (Earnest Insight Toolkit), a tool designed to assess students' engagement within interactive lecture participation exercises-particularly in the context of large-scale hybrid classrooms. We use EIT to conduct a comprehensive assessment of student responses to interactive lecture poll questions. Our objective with EIT is to equip educators with valuable means of identifying at-risk students for enhancing intervention and support strategies and measuring student engagement with course content.

In this work, we present the development of an automated extension tool that supports flexible extension policies. Students interact with a wide range of extension policies in similar ways; in particular, some students repeatedly request multi-day long extensions. When scaled to courses with hundreds or potentially thousands of students, course staff time is the limiting resource preventing adequate student support. We present a tool to help automate a range of extension processes. The use of this tool should reduce staff load while increasing individualized student support, through email communication and consequent recovery of student agency. Our early research questions are: Does the extension tool reduce barriers and stigma around asking for assistance? Does the tool lessen the wait time between requesting and receiving an extension, and how does the tool improve students' learning experience in the course? These questions will help inform us about how an automated tool for flexible extensions helps support growing course sizes and students who may not otherwise receive the support they need for their success and well-being in the course.

Equitable grading practices and flexible deadline policies have previously demonstrated positive student learning and well-being outcomes. In this poster, we contribute a framework for flexible extension policies that emphasize equitable grading. We then analyze extension requests and grades obtained by students in a Data Science course with a flexible extension policy. We present two research questions based on this data. RQ1: How does the length of an extension relate to student performance on the corresponding assignment? RQ2: How does student extension usage across the semester relate to students' learning of the content?

The Early Research Scholars Program (ERSP) was launched in 2014 at UC San Diego as a way to provide the benefits of research experiences to a large and diverse group of students early in their undergraduate computing career. ERSP is a structured program in which second-year undergraduate computing majors participate in a group-based, dual-mentored research apprenticeship over a full academic year. In its first four years ERSP engaged 139 students with a high proportion of women (68%) and racially minoritized students (19%), and participation in ERSP correlated with increased class grades. In 2018 we partnered with three additional universities to launch their own version of ERSP. Implementations at our partner sites have seen similar diversity and initial success, and have taught us how to implement the program in different contexts (e.g. quarters vs. semesters, different credit structures). This paper describes the structure of ERSP and how it can be adapted to different contexts to construct a scalable and inclusive research experience for early-career undergraduates in computing and related fields.

During the COVID-19 pandemic, many educators have had to rethink their methodology for summative assessment. Are timed and proctored exams appropriate---or even feasible---in this new open-internet, online learning environment? In this experience paper, we discuss our unique exams framework: our tool for generating exams that are uniquely identifiable but conceptually identical. In our university-level Probability for Computer Scientists Course, students completed unique exams generated from a common exam skeleton, with unique numeric variations per problem. With few deviations from the creation, administration, and grading processes of a traditional exam, our framework can provide a layer of security for both students and instructors about exam reliability for any classroom environment---in-person or online. In addition to sharing our experience designing unique exams, in this paper we also present a simple end-to-end tool and example question templates for different CS subjects that other instructors can adapt to their own courses.

Programming is fast becoming a required skill set for students in every country. We present CS Bridge, a model for cross-border co-teaching of CS1, along with a corresponding open-source course-in-a-box curriculum made for easy localization. In the CS Bridge model, instructors and student-teachers from different countries come together to teach a short, stand-alone CS1 course to hundreds of local high school students. The corresponding open-source curriculum has been specifically designed to be easily adapted to a wide variety of local teaching practices, languages, and cultures.

In large undergraduate computer science classrooms, student learning on assignments is often gauged only by the work on their final solution, not by their programming process. As a consequence, teachers are unable to give detailed feedback on how students implement programming methodology, and novice students often lack a metacognitive understanding of how they learn. We introduce Pensieve as a drag-and-drop, open-source tool that organizes snapshots of student code as they progress through an assignment. The tool is designed to encourage sit-down conversations between student and teacher about the programming process. The easy visualization of code evolution over time facilitates the discussion of intermediate work and progress towards learning goals, both of which would otherwise be unapparent from a single final submission. This paper discusses the pedagogical foundations and technical details of Pensieve and describes results from a particular 207-student classroom deployment, suggesting that the tool has meaningful impacts on education for both the student and the teacher.

In the ideal CS1 classroom, we should understand programming process---how student code evolves over time. However, for graphics-based programming assignments, the task of understanding and grading final solutions, let alone thousands of intermediate steps, is incredibly labor-intensive. In this work, we present a challenge, a dataset, and a promising first solution to autonomously use image output to identify functional, intermediate stages of a student solution. By using computer vision techniques to associate visual output of intermediate student code with functional progress, we supplement a lot of the teacher labor associated with understanding graphics-based, open-ended assignments. We hope our publication of the dataset used in our case study sparks discussion in the community on how to analyze programs with visual program output.

As computer science classes grow, instructor workload also increases: teachers must simultaneously teach material, provide assignment feedback, and monitor student progress. At scale, it is hard to know which students need extra help, and as a result some students can resort to excessive collaboration--using online resources or peer code--to complete their work. In this paper, we present TMOSS, a tool that analyzes the intermediate steps a student takes to complete a programming assignment. We find that for three separate course offerings, TMOSS is almost twice as effective as traditional software similarity detectors in identifying the number of students who exhibit excessive collaboration. We also find that such students spend significantly less time on their assignment, use fewer class tutoring resources, and perform worse on exams than their peers. Finally, we provide a theory of the parametric distribution of typical student assignment similarity, which allows for probabilistic interpretation.

As datacenter speeds scale to 100 Gb/s and beyond, traditional congestion control algorithms like TCP and RCP converge slowly to steady sending rates, which leads to poorer and less predictable user performance. These reactive algorithms use congestion signals to perform gradient descent to approach ideal sending rates, causing poor convergence times. In this paper, we propose a proactive congestion control algorithm called PERC, which explicitly computes rates independently of congestion signals in a decentralized fashion. Inspired by message-passing algorithms with traction in other fields (e.g., modern Low Density Parity Check decoding algorithms), PERC improves convergence times by a factor of 7 compared to reactive explicit rate control protocols such as RCP. This fast convergence reduces tail flow completion time (FCT) significantly in high speed networks; for example, simulations of a realistic workloads in a 100 Gb/s network show that PERC achieves up to 4x lower 99th percentile FCT compared to RCP.

We propose, analyze and implement a general architecture for massively parallel VoD content distribution. We allow for devices that have a wide range of reliability, storage and bandwidth constraints. Each device can act as a cache for other devices and can also communicate with a central server. Some devices may be dedicated caches with no co-located users. Our goal is to allow each user device to be able to stream any movie from a large catalog, while minimizing the load of the central server. First, we architect and formulate a static optimization problem that accounts for various network bandwidth and storage capacity constraints, as well as the maximum number of network connections for each device. Not surprisingly this formulation is NP-hard. We then use a Markov approximation technique in a primal-dual framework to devise a highly distributed algorithm which is provably close to the optimal. Next we test the practical effectiveness of the distributed algorithm in several ways. We demonstrate remarkable robustness to system scale and changes in demand, user churn, network failure and node failures via a packet level simulation of the system. Finally, we describe our results from numerous experiments on a full implementation of the system with 60 caches and 120 users on 20 Amazon EC2 instances. In addition to corroborating our analytical and simulation-based findings, the implementation allows us to examine various system-level tradeoffs. Examples of this include: (i) the split between server to cache and cache to device traffic, (ii) the tradeoff between cache update intervals and the time taken for the system to adjust to changes in demand, and (iii) the tradeoff between the rate of virtual topology updates and convergence. These insights give us the confidence to claim that a much larger system on the scale of hundreds of thousands of highly heterogeneous nodes would perform as well as our current implementation.

To maximise the assignment of function of the proteins encoded by a genome and to aid the search for novel drug targets, there is an emerging need for sensitive methods of predicting protein function on a genome-wide basis. GeneAtlas is an automated, high-throughput pipeline for the prediction of protein structure and function using sequence similarity detection, homology modelling and fold recognition methods. GeneAtlas is described in detail here. To test GeneAtlas, a 'virtual' genome was used, a subset of PDB structures from the SCOP database, in which the functional relationships are known. GeneAtlas detects additional relationships by building 3D models in comparison with the sequence searching method PSI-BLAST. Functionally related proteins with sequence identity below the twilight zone can be recognised correctly.