This research-to-practice full paper addresses the important need for relevant and comprehensive datasets to advance cybersecurity research by proposing methods for curating open datasets and generating synthetic datasets. Cybersecurity research is a rapidly evolving scientific field, making robust datasets crucial for empirical analysis. Unfortunately, current doctoral research is hindered by the scarcity, limited accessibility, and outdated or irrelevant nature of existing open-source datasets. This paper tackles these challenges by focusing on two main initiatives: (1) curating a pilot collection of open datasets aligned with the National Initiative for Cybersecurity Education (NICE) Cybersecurity Workforce Framework, and (2) using Generative Adversarial Networks (GANs) to generate synthetic datasets. Our research highlights the obstacles faced by doctoral students due to fragmented, outdated data and underscores the importance of accessible datasets for rigorous scientific inquiry. We also demonstrate how synthetic data can ease privacy concerns while still offering researchers realistic data. By incorporating these approaches into doctoral curricula, we aim to equip future cybersecurity researchers with the skills resources for impactful research. The authors will continue to expand their dataset curation efforts and study how discoverable, high-quality datasets can influence doctoral research, particularly empirical studies and their outcomes.

This research explores the use of virtual reality (VR) in Computer Science education to enhance the visualization and understanding of algorithm complexity and data structures. The proposed VR system aims to provide undergraduate students with a more tangible and interactive learning experience, allowing them to interact with algorithms and data structures in a virtual environment. Through the implementation of immersive techniques and real-time visualization, students can manipulate and observe the behavior of these abstract concepts. The VR application covers topics such as algorithm analysis, queues, stacks, tree structures, and algorithmic complexity. The use of VR in Computer Science education has the potential to transform teaching methods and provide instructors with valuable supplementary materials. Future studies will evaluate the impact of VR on student learning outcomes, and the VR application will be made available for further research and instruction in the field.

Many social robots have been developed to support the needs of users, such as tour guides [1] or sales robots [2], [3]. In these systems, the main purpose of the human robot interaction is to support the user's need. However, what if in addition to these capabilities, the robot had a goal of persuading the user to do something of which the user had no knowledge. What would the user's perceptions on the interaction be? We developed a social robot with the ability to employ six types of persuasion conversation logic, namely, scarcity, emotion, social identity, commitment, concreteness, and no persuasion [4] and measured the users' attitudinal and behavioral responses when interacting with our robot. In this pilot study we describe our initial results with success rates varying across all six persuasion techniques. Particular persuasion techniques demonstrated as high as a 75% success rate at directing users towards a secret task.

Co-clustering is the problem of deriving sub-matrices from the larger data matrix by simultaneously clustering rows and columns of the data matrix. Traditional co-clustering techniques are inapplicable to problems where the relationship between the instances (rows) and features (columns) evolve over time. Not only is it important for the clustering algorithm to adapt to the recent changes in the evolving data, but it also needs to take the historical relationship between the instances and features into consideration. We present ESCC, a general framework for evolutionary spectral co-clustering. We are able to efficiently co-cluster evolving data by incorporation of historical clustering results. Under the proposed framework, we present two approaches, Respect To the Current (RTC), and Respect To Historical (RTH). The two approaches differ in the way the historical cost is computed. In RTC, the present clustering quality is of most importance and historical cost is calculated with only one previous time-step. RTH, on the other hand, attempts to keep instances and features tied to the same clusters between time-steps. Extensive experiments performed on synthetic and real world data, demonstrate the effectiveness of the approach.