Computer Science and Business
Modern CI/CD Pipelines for Frontend and Backend Web Applications
Heriberto Luna
Florida International University
A continuous integration pipeline is an automated set of steps to ensure code is of high quality and properly tested before it is integrated into a repository. All these steps are tedious to do manually. Continuous deployment shares some similarities with CI. Whenever code is added to a repository and built, the continuous deployment pipeline can be executed to automatically deploy code into production and monitor it. Why is CI/CD important? Because the more work that is automated, the less that must be done, and this facilitates the maintenance and management of a project. This paper will explore two implementations of a CI/CD pipeline for both the frontend and backend of a full-stack application. That is why the area being explored throughout this paper is CI/CD. Thus, the following questions can be asked: "How to design a proper CI/CD pipeline for frontend and backend web development?" and "How can AI be integrated into existing CI/CD pipelines?" The CI/CD pipeline will increase code quality, the rate of development, and the deployment of services. It can be concluded that CI/CD pipelines are useful for automating tasks, reducing development time, and increasing the quality of projects. Future research regarding this topic may include further use of AI for automation, the testing of other CI tools such as Jenkins, and the development of tools for the creation of CI/CD pipelines. I would like to acknowledge the CIARA research center and Dr. Vasilka Chergarova for their support in this research.
A Deep-Agent Cyber Investigation System Using Event-Centric Graph Learning for Red-Team Detection and Lateral-Movement Analysis
Francisco Barraza Segura
Our Lady of the Lake University
Enterprise networks can generate millions of authentication events each day where the traces of an intruder's lateral movement can be easily lost among them. This research developed a deep-agent cyber investigation system for detecting red-team activity and analyzing possible lateral movement in the LANL cyber dataset. The system models cyber logs as an event-centric heterogeneous graph in which authentication events, process events, users, computers, processes, network-flow activity, and DNS relationships are preserved as connected graph objects. The framework operates in three stages. Because malicious events are rare when compared to benign ones, we treat this operation as a ranking problem. For the first stage, we use machine-learning and graph–learning models to rank authentication events by their likelihood of being red team activity. The second stage uses high-scoring authentication events to score possible lateral-movement paths through users, computers, process events, and network connections. Third, a deep-agent investigation system pairs a modern agent harness with tools trained on graph-learning and a GraphRAG tool. over a well-structured relational representation, letting the agent investigate far more effectively than reasoning over raw logs. It uses the ML alerts, GNN scores, lateral-movement path scores, and graph evidence as tools to investigate suspicious activity and make a final determination about whether the activity is malicious or not. The complete system is evaluated both at the model level and end-to-end using detection metrics such as precision, recall, F1-score, AUPRC, and analyst-style alert quality. This framework can automate cybersecurity protocols and benefit networks of all sizes.
Mapping Communication Expectations Across the Business Core: A Textbook-Based Genre and Values Analysis
Kaden Huff
Baylor University
Business communication courses are central to students’ professional readiness but can complicate learning application and writing expectations. Limited conversation between business communication faculty and colleagues in other business disciplines can create siloed instruction, which results in students encountering varied and conflicting standards for writing and presentation expectations. Our research aims to develop a clearer, research-grounded picture of disciplinary genres, values, and evaluative criteria across the Baylor University Hankamer Business School’s core thirteen classes. We conducted a systematic analysis of recent textbooks assigned in our university’s Bachelor of Business Administration core courses. We coded textbooks for recurring communication expectations using a structured set of dimensions informed by genre studies, transfer research, VALUE rubrics, and business disciplinary communication scholarship (AAC&U n.d.; Adler-Kassner et al., 2017). Emerging findings suggest indicates some similarities in communication genre, values, and criteria across core business courses. Findings also suggest differences in these areas, supporting prior studies that business writing and speaking tasks vary across subfields in their purposes, evidence norms, and rhetorical expectations (Zhu, 2004). Implications include proposing rubrics that adapt VALUE-style dimensions to business competencies to align with U.S. business school standards (AAC&U, n.d.; AACSB, 2021).