Software Technology

The expanding capabilities of Generative AI are forging new paths in software development, an area ripe with opportunities for innovation and transformation. This minitrack aims to encapsulate a comprehensive view of this evolving landscape, highlighting both the technological advancements and the broader implications of these emergent technologies.

We seek to attract a cadre of research that both delineates and critiques the concepts, methods, frameworks, architectures, functionalities, and broader implications of applying and integrating Generative AI in software development. The scope of this minitrack includes, but is not limited to, the following key areas:

• Deployment of synthetic data for model training within development environments, facilitating advanced machine learning applications while addressing data privacy concerns.
• Exploration of the ethical dimensions in the use of Generative AI in software development, focusing on intellectual property rights, bias in automated outputs, and accountability in decision-making.
• Integration of Generative AI into coding practices, improving code quality and development speed, while considering issues of code originality and skill enhancement.
• Utilization of Generative AI for creating comprehensive test datasets, enhancing software testing effectiveness, and uncovering critical edge cases.
• Intersection with agile and DevOps methodologies, leveraging Generative AI for continuous integration and adaptive responses to evolving project requirements.
• Implications of Generative AI in legacy system sustainment, addressing the challenges and opportunities in modernizing and maintaining older software infrastructures.
• The role and impact of Generative AI in software innovation, fostering new methods in ideation, brainstorming and co-designing, thus revolutionizing traditional approaches to software development and project conceptualization.

This minitrack aspires to be a platform for rigorous scholarly inquiry into the multifaceted applications of Generative AI in software development, emphasizing both the innovative potential and the consequential ethical, legal, and operational challenges.

Minitrack Co-Chairs:

Johnny Chan (Primary Contact)
University of Auckland
jh.chan@auckland.ac.nz

Gabrielle Peko
University of Auckland
g.peko@auckland.ac.nz

David Sundaram
University of Auckland
d.sundaram@auckland.ac.nz

Ghazwan Hassna
Hawaiʻi Pacific University
ghassna@hpu.edu

This minitrack intends to serve as a platform for presenting innovative research, discussing emerging trends, and setting the agenda for future exploration in the field of AI security. It will address the intersection of technology, safety, trustworthiness, responsibility, and policy, ensuring a comprehensive approach to AI security. The focus will be placed on developing and implementing mechanisms to ensure that AI systems are safe for users and operate in a trustworthy manner. This includes research on creating transparent AI systems that users can understand and trust, and methodologies to verify and validate the safety of AI algorithms in various applications. Recommended topics include, but are not limited to, the following:

• System-Theoretic Process Analysis for Security (STPA-SEC) for AI Systems
• AI Safety and Trustworthiness
• AI Responsibility
• AI Risk Models
• AI Explainability
• AI Ethics and Algorithmic Bias
• AI Cybersecurity, Offensive, and Defensive Operations
• Secure Machine Learning Operations (MLOps)
• Secure AI Algorithm and Machine Learning Algorithms
• Red Teaming AI systems and algorithms
• Robustness and Resilience of AI Systems
• Threat Detection and Response in AI Systems
• Ethical and Regulatory Considerations in AI Security

There will be opportunities for researchers to fast track a journal publication in a Special Issue within the Proceedings of the IEEE.

Minitrack Co-Chairs:

Tyson Brooks (Primary Contact)
Department of Defense and Syracuse University
ttbrooks@syr.edu

Shiu-Kai Chin
Syracuse University
skchin@syr.edu

William Young
Syracuse University
weyoung@syr.edu

Erich Devendorf
Air Force Research Laboratory
erich.devendorf.1@us.af.mil

Game engines have evolved beyond their traditional use in video game development and are now utilized in various application areas. Important examples of such application areas include film and animation production, architectural visualization and design, media arts, digital performance and theatre, digital health and well-being, training and simulation, data visualization, mixed reality, ar/vr, serious games and gamification, education and e-learning, interactive exhibits and installations, and non-technical aspects of game technology and game experience.

This minitrack is open for anybody working with gaming technology from implementation to content creation with a particular focus on applications beyond game development. Topics of interest include (but are not limited to):

• New applications of game technology (e.g., virtual production)
• End user programming with game engines
• Extending game engines for new application areas
• User experience design based on game technology
• Content creation workflows
• Teaching game technology
• Experience reports from practical applications

Minitrack Co-Chairs:

Tim Schattkowsky (Primary Contact)
Hamm-Lippstadt University of Applied Sciences
tim.schattkowsky@hshl.de

Christian Geiger
University of Applied Sciences Düsseldorf
geiger@hs-duesseldorf.de

Stefan Albertz
Hamm-Lippstadt University of Applied Sciences
stefan.albertz@hshl.de

The societal importance of software systems and cyber-physical systems is increasing as they affect almost every aspect of our daily lives. With applications ranging from digital twins of individuals to entire societies or individual industrial processes to autonomous vehicle fleets, the impact of these systems has reached a new level where they can transform society and the daily lives of individuals. This presents various new and greater challenges that need to be considered from an interdisciplinary perspective. They include:

• Systematic development of digital twins
• Federated, cooperated digital twins
• Architectural and design patterns
• Privacy concerns, security risks, and safety implications
• Increasing resilience, availability, and dependability of software systems
• Environmental impact by digital twins on software systems
• Regulations and regulatory compliance

Addressing these challenges requires collaboration among stakeholders from various domains, including science, technology, policy-making, and civil society, to develop comprehensive solutions that prioritize societal well-being and promote responsible technology usage.

This minitrack intends bring together these stakeholders to present research and practical experience from all related areas as a basis for an interdisciplinary discussion. Topics of interest include:

• Technologies, frameworks, platforms, and architectures for modern software systems
• Case studies of socio-technical ecosystems with federated, cooperating digital twins
• Investigating the social impact of software- and cyber-physical systems
• Incorporating social awareness in software engineering methods

Minitrack Co-Chairs:

Achim Rettberg (Primary Contact)
Carl-von-Ossietzky Universität Oldenburg
achim.rettberg@uni-oldenburg.de

Gregor Engels
Paderborn University
engels@upb.de

Stefan Henkler
Hamm-Lippstadt University of Applied Sciences
stefan.henkler@hshl.de

Tim Schattkowsky
Hamm-Lippstadt University of Applied Sciences
tim.schattkowsky@hshl.de

As technology is incorporated into more aspects of daily life, cyber operations, defenses, and digital forensics solutions continue to evolve and diversify. This encourages the development of innovative managerial, technological, and strategic solutions. Hence, a variety of responses are needed to address the resulting concerns. There is a need to research a) technology investigations, b) technical integration and solution impact, c) the abuse of technology through attacks, and d) the effective analysis and evaluation of proposed solutions. Identifying and validating technical solutions to secure data from new and emerging technologies, investigating these solutions’ impact on the industry, and understanding how technologies can be abused are crucial to the viability of commercial, government, and legal communities.

We welcome new, original ideas from academia, industry, government, and law enforcement participants interested in sharing their results, knowledge, and experience. Topics of interest include but are not limited to:

• Detection and analysis of advanced threat tactics, techniques, and procedures
• Applying machine learning tools and techniques in terms of cyber operations, defenses, and forensics
• Case studies surrounding the application of policy in terms of cyber operations, defenses, and forensics
• Approaches related to threat detection and Advanced Persistent Threats (APTs)
• Solutions that secure different types of data stored in different layers of computer networks
• “Big Data” solutions and investigations – collection, analysis, and visualization of “Big Data” related to cyber operations, defenses, and forensics
• Malware analysis and the investigation of targeted attacks
• Device investigations that assist with the recovery and reconstruction of digital artifacts
• Digital evidence recovery, storage, preservation, and memory analysis
• Event reconstruction approaches and techniques
• Anti-forensics techniques and solutions
• Investigations related to mobile devices, embedded systems, or Internet of Things (IoT) devices
• Forensic investigations within emerging domains such as transportation systems, industrial control systems, and SCADA
• Network investigations – collection, analysis, and visualization of network forensic data
• Privacy implications related to security incident response and digital forensic investigations
• Research in security incident management
• Situational awareness related to security incident response
• The impact of digital evidence on the legal system

The above list is suggestive. Authors are encouraged to contact the minitrack chairs to discuss related topics and their suitability for submission to this minitrack.

Accepted papers will be offered the opportunity to extend their submission by 50% and submit to a special issue of the Association for Computing Machinery (ACM) Digital Threats: Research and Practice (DTRAP) Journal.

Minitrack Co-Chairs:

William Bradley Glisson (Primary Contact)
Louisiana Tech University
glisson@latech.edu

Todd McDonald
University of South Alabama
jtmcdonald@southalabama.edu

Modern society is irreversibly dependent on software systems of astonishing scope and complexity. Yet despite best efforts, errors, vulnerabilities, failures, and compromises continue to persist. Networked systems with complex hardware and software components embody many pathways that adversaries can exploit. Experience shows that contemporary cybersecurity and software assurance methods are insufficient to meet this challenge.

This minitrack focuses on how to enable development and application of these foundations. We ask: How should research and development move us toward a solid basis in understanding and principle? The goal is to develop science foundations, technologies, and practices that can improve the security and dependability of complex systems. This minitrack will bring together researchers in cybersecurity and software assurance in a multidisciplinary approach to these problems. It addresses all aspects of cybersecurity viewed from an “enlightened security” perspective. The following topics will be included in the minitrack:

• Security ecosystem
• Designed-in security
• Tailored trustworthy spaces
• Moving target
• Cyber economics
• Science of security
• Multivariate detection and response
• Co-evolution of defense and offense
• Biologically-inspired security models
• Holistic risk analysis
• Hardware-enabled trust
• Layered adaptable defense
• Real-time coordinated response
• Automated system interoperability
• Authentication in ecosystem
• Practical use of continuous monitoring
• Confidence in activity prediction
• Security visualization and prediction
• Theories of vulnerability classification and control
• Security measurement
• • Advances in information assurance theory and practice
• Advances in specification, design, and implementation of assured systems
• Advances in verification, testing, and certification of assured systems
• Advances in software security analysis
• Assurance for embedded systems and hardware components
• Assurance for large-scale infrastructure systems
• Information and software assurance in cloud computing environments
• Assurance in system maintenance and evolution
• Automated methods for information and software assurance
• Assurance through computation of software behavior
• Management of assurance operations
• Processes and metrics for information and software assurance
• Business case and ROI development for information and software assurance
• Supply chain and standards issues in information and software assurance
• Case studies of system assurance successes
• Software testing

Minitrack Co-Chairs:

Luanne Burns Chamberlain (Primary Contact)
Johns Hopkins University Applied Physics Lab
luanne.chamberlain@jhuapl.edu

Thomas Llanso
John Hopkins University Applied Physics Lab
thomas.llanso@jhuapl.edu

Richard George
Johns Hopkins University Applied Physics Lab
richard.george@jhuapl.edu

Games for Impact minitrack intends to draw attention to the use of games and game technology for special purposes and positive outcomes where the created experience reaches beyond entertainment. Recognizing that games are a powerful vehicle to make emerging technologies accessible to society, this mini track creates a space to explore the many factors that influence the design, development, application, adoption, use, and impact of games and game technology.

The exploration of Games for Impact mini track falls under the umbrella of recent games research fields such as games for health/rehabilitation/therapy, games for learning, games for empathy, games for social innovation, and citizen science games. Potential subtopics or areas including but not limited to the following are listed below:

• Case study on designing, developing, using, and evaluating games for special purposes
• Best practices and guidelines on game design, study design, interaction design, user experience (UX) and user interface (UI)
• The role and application of games and game technology in creating, disseminating, and evaluating social innovation
• The application and impact of games and game technology in education and its accessibility
• The application and impact of games for training, learning, and personal development (habit building, empathy, social skills, etc.)
• Evaluation approaches, criticality, quality measures, and ethics of using and adapting games and games technology in other fields such as health, rehabilitation, education, social innovation, citizen science
• Use of novel interaction modalities, platforms and/or controllers, IoT, VR-AR-MR
• Analysis for socio-cultural context of games for impact
• Demographics, persona studies, and ethics of the application, adoption, and impact of games and game technology for purposes beyond entertainment

We welcome contributions on design and development methods, technical studies that focus on implementation and development guidelines, case studies with novel interaction modalities including platforms (mobile, AR-VR-MR) and/or controllers, user experience (UX) approaches, user interface (UI) techniques, analysis for socio-cultural context of games for impact, demographics and persona studies, and ethical studies for the aforementioned research fields.

Minitrack Co-Chair:

This minitrack aims to explore the intersection of AI and software engineering, focusing on the innovative ways in which AI technologies are influencing software development, testing, maintenance, and overall software lifecycle management. It invites researchers and practitioners to delve into the multifaceted implications of AI on software engineering practices, providing a platform for insightful discussions and the exchange of cutting-edge research findings.Topics of Interest include, but are not limited to:

• AI-driven Software Development Processes:
  • Automated code generation and optimization
  • Impact on Software Design
  • Intelligent code completion and suggestion systems
  • AI-assisted requirement analysis and specification
• AI in Software Testing and Quality Assurance:
  • Automated testing using machine learning algorithms
  • AI-driven fault prediction and localization
  • Quality assurance in AI-infused software systems
• AI for Software Maintenance and Evolution:
  • Predictive maintenance and malfunction detection
  • Intelligent bug tracking and resolution
  • Adaptive software evolution with AI assistance
• Ethical and Social Implications of AI in Software Engineering:
  • Bias and fairness in AI-enhanced software systems
  • Responsible AI practices in software development
  • Societal impact of AI-driven software solutions
  • How are the roles reshaped in software development teams
• Educational Initiatives in AI and Software Engineering:
  • Integration of AI concepts into software engineering curricula
  • Training programs for AI-aware software engineers
  • Challenges and opportunities in AI education for software developers

Minitrack Co-Chairs:

Stefan Wittek (Primary Contact)
Clausthal University of Technology
switt@tu-clausthal.de

Sandra Gesing
San Diego Supercomputing Center
sgesing@ucsd.edu

Peter Salhofer
FH JOANNEUM
peter.salhofer@fh-joanneum.at

We would like to invite papers which focus on developing paradigms for addressing the synergy of software, devices, moveable objects, networks, and people, in one constantly changeable environment.  We would like to debate challenges of empowering edges of computer networks and misconceptions on their capabilities of carrying out modern computations. Examples of dynamic edge computing with the Internet of Things (IoT) / Internet of Everything (IoE) are more than welcome.  The IoT/IoE provide the possibility of constantly changeable sets of new types of connected devices, which might be wearables, belong to soft robotics, represent computational materials, and enable human machine augmentations.  In all these examples, edge computing plays an important role, and it is no surprise that we have started thinking how to equip the edges with intelligence.

However, there is something more important on the horizon; the latest advances in predictive and learning software technologies, such as generative AI models, are democratizing AI and leading public fascination with the current AI and this is pushing forward ideas of introducing generative AI at computational edges. Given the computational demands of generative AI,  emphasis is needed on sustainability, energy efficiency and adaptability of edge solutions, bearing in mind that we may have to descale and distribute current generative AI models and their data sources, and thus rethink the role of generated contents suitable for edge computing.  There are other challenges in promoting intelligent edge.  Most of them resulting from the uncertain future of replacing powerful cloud computing with new paradigms in which intelligent edge computing may take centre stage. Topics of interests include:

• Towards Edge Computing Paradigm
  • Assessing computational edge technology landscape and its potentials for creating computational paradigms for the edge
  • Analysis of computability within edge technology landscape: the power of data, scalability and sustainability of computations
  • Examples of computing paradigms suitable for localised computing and its extensions towards Fog/Cloudlets/Clouds
  • Pushing data and computational intelligence, with analytic platforms, away from centralised Clouds and the role of serverless computing
  • The current role of Edge/Fog computing in IoT applications and implementations: reconsidering the role of Clouds
• Creating Intelligent Edge Computing
  • Roadmap for AI at the computational edge, challenges, and opportunities
  • Feasibility of creating intelligence “at the edge” which could share computations and data between Clouds, Fog and Cloudlets and accommodate predictive technologies
  • Software architectures for supporting intelligence at the computational edge for the purpose of creating solutions in automation, manufacturing, businesses, medicine, healthcare delivery, education, and governance
  • Edge computing hardware solutions, including neuromorphics, tensor processing units (TPUs), FPGAs, and other dynamically programmable edge devices
  • Potential convergence of humans,” things” and AI in creating edge intelligence
  • Possibilities of defining a computational continuum by utilizing the space between localized computing and clouds and thus enhance the intelligence at the edge
• Generative AI and Sustainable Edge Computing
• • The power of edge computing when addressing sustainability of running applications of generative AI
  • Leveraging edge computing and generative AI and tuning LLM at computational edges for specific tasks
  • Deploying generative AI inference at the edge
  • Optimizing edge performance for specific generative AI tasks, in certain contexts with specific datasets close to edges
  • Software architectures for managing distributions and tuning of and inference in generative AI models at the edge
  • Developing and tuning of smaller Language Models at the edge, for specific tasks
  • Edge computing hardware solutions for performing generative AI tasks
  • Analyzing Investments of hardware manufacturers in generative AI and energy sustainable edge devices, including neural and tensor processors
• Edge Computing for/in Human Augmentation, Internet of Materials and Human centered Cognitive IoT
  • The role of edge computing in creating augmentation of human with machines and vice versa.
  • Promoting intelligent engineering for designing cognitive devices ready to accommodate edge intelligence
  • Investigating computational materials, shape changing computing particles and internet of materials as potential factors for empowering intelligent edge
  • Pushing forward the fusion of computational materials with hardware/software synergy typical of binary/ternary/quantum computers

Minitrack Co-Chairs:

Radmila Juric (Primary Contact)
ALMAIS Consultancy
radjur3@gmail.com

Elisabetta Ronchieri
INFN-CNAF Bologna and University of Bologna
elisabetta.ronchieri@cnaf.infn.it

Filippo Sanfilippo
University of Agder
filippo.sanfilippo@uia.no

Trevor Bihl
Air Force Research Laboratory
Trevor.bihl.2@us.af.mil

Minitrack Co-Chairs:

William Yeager (Primary Contact)
Stanford University Knowledge Systems Lab, Retired
byeager@fastmail.fm

Jean-Henry Morin
University of Geneva
Jean-Henry.Morin@unige.ch

This minitrack is dedicated to exploring the critical intersection of software usability, sustainability, and reproducibility, acknowledging the expanding role of software in shaping research across diverse domains. The three concepts usability, sustainability and reproducibility are interconnected with each other and cover a wide range of application areas. They affect all layers of the software process – from enabling reproducing experiments via an easy user interface to using containerization for application portability. Such concepts are also relevant in the building of Science Gateways (also known as virtual laboratories or virtual research environments), which by definition serve communities with end-to-end solutions tailored specifically to their needs. The mini-track will focus on the broad spectrum of submissions that deal with complex scenarios such as containerization, strategies for long-lasting software, usability and user interface issues, handling data curation and provenance and more.

Topics of interest include but are not limited to:

• Web-based solutions (web sites, science gateways, virtual labs, etc.)
• Application Programming Interfaces (APIs)
• Computational and Data-Intensive Workflows
• Novel approaches in containerization
• Sustainability practices in software development
• System architectures for testing and continuous integration
• Emerging best practices in Machine Learning software
• Best practices and Key Success Factors for usability, sustainability and reproducibility
• Community building practices
• Sustainability practices in software development, with a focus on AI applications
• System architectures for testing and continuous integration in AI systems
• Emerging best practices in AI and Machine Learning software
• Addressing ethical considerations in AI-related software
• Best practices and Key Success Factors for usability, sustainability, and reproducibility in the context of AI

Minitrack Co-Chairs:

Maytal Dahan (Primary Contact)
University of Texas at Austin
maytal@tacc.utexas.edu

Joe Stubbs
University of Texas at Austin
jstubbs@tacc.utexas.edu

Sandra Gesing
San Diego Supercomputing Center
sgesing@ucsd.edu

The development of software has provided ample opportunities for research, provides ample oppor-tunities, and likely will provide ample opportunities. Not long ago, the proliferation of mobile computing opened up a new stream of research, then the same happened with the Internet-of-Things (IoT) and Cyber-Physical Systems (CPS). Fog, edge, and dew computing and the convergence of technologies likely will continue this trend. All these topics seemingly provide completely new endeavors; with a closer look, however, they can draw from what is already known – both regarding typical problems and regarding solutions.

Experiences and methods from classical software development can only be utilized to some degree when solving challenges that arise from new application, changing environment, and demanding do-mains. Development is complicated by the often faced need to develop for a multitude of platforms. With the emergence of multi-platform and multi-device, the new golden standard are applications not only across software ecosystems, but across hardware platforms such as laptop, mobile, tablets, em-bedded devices, sensors, and wearables. Therefore, new threads of research are needed to tackle these issues and to pave the way for improved software development, better business producibility and improved user experience (UX).

Further, there are novel developments in machine learning and analysis, and the emergence of multi-faceted aspects of artificial intelligence (AI), ranging from algorithms to ethical AI, secure AI and sustainable AI perspectives. This creates new opportunities for groundbreaking research through distributed machine learning, federated learning, edge analytics and computational collaboration between several heterogenous systems and device forms.

The minitrack is devoted to the technological background while keeping an eye on business value, user experience, and domain-specific issues. Contributions may take a socio-technical view or report on technological progress. We are particularly interested in applied software technology but also welcome theoretical work. Topics of interests include, but not limited to:

• Case studies of development
• Development methods, software architectures, and specification techniques
• Economic and social impact, behavioral aspects
• Software engineering education
• User interface (UI) design and user experience (UX) research
• Hybrid and cross-platform development
• Web technology
• Security, safety, and privacy
• Energy-efficient computing
• Machine learning on device
• The convergence between mobile devices, IoT, and CPS
• Fog, edge, and dew computing and their computational applications

Minitrack Co-Chairs:

Tim A. Majchrzak (Primary Contact)
University of Agder
tim.a.majchrzak@uia.no

Tor-Morten Grønli
Kristiania University College
tor-morten.gronli@kristiania.no

Hermann Kaindl
TU Wien
kaindl@ict.tuwien.ac.at

New approaches, methods, and tools (some of which are AI-enabled) for facilitating more responsive organizations are proliferating rapidly. Among the evolving methods are approaches such as Agile, lean, DevOps, BizDevOps. In addition, low-code and co-pilot platforms are being increasingly used to accelerate software development but can create organizational challenges.

The organizational challenges are numerous. For example, Agile was initially designed for colocated, onsite teams, but organizations today cope with scaling issues and remote and hybrid work. Low code software development enables non-software development personnel to create applications, but those personnel may lack sufficient knowledge of good software development practices. The challenges with artificial intelligence-developed code are similar, but perhaps even more extreme. Lean business models assume collocated access to the customer, but often startups now serve geographically dispersed customers.

In this minitrack, we seek research papers and experience reports that explore practices, tools, and techniques for rapid development. We also seek to explore how these concepts can be leveraged in other contexts (such as data science or physical product development). Practitioners interested in submitting an experience are welcome to reach out to a minitrack co-chair for support and guidance, if desired.

Our minitrack seeks to answer questions such as:

• How can emerging technologies like AI and machine learning be seamlessly integrated into existing software development practices to enhance efficiency and effectiveness?
• How to balance team autonomy and decentralized decision-making with the need for organizational control and alignment in large-scale agile development?
• How can agile and lean can be integrated within a single coherent approach?
• Which metrics help enterprises, teams and individuals adapt and improve? What common behaviors do we see in agile or lean teams and how do those behaviors affect outcomes?
• How do organizations implement, monitor and improve hiring, coaching, training and mentoring?
• How to scale agile (how to effectively manage dependencies, teams, stakeholders, processes, technologies, and tools)including comparative results on the use of different agile scaling frameworks?
• How can agile be implemented within other contexts (e.g., data science,BizDevOps)?
• What organizational structures are required to enable shared leadership in self-managed teams?
• How to balance the need for effective coordination and focused work in an agile team?
• How do agile and lean principles extend to DevOps environments? Is there a difference between agile and lean before and after deployment? How are post-deployment issues and opportunities in software projects impacting planning and development of software development projects?
• What organizational structures and novel tools are required to leverage AI, low-coding, and rapid-prototyping as part of the project management process?
• What are the best practices for maintaining efficiency and effectiveness in remote or hybrid agile teams?
• How can agile teams ensure inclusivity and leverage diversity to enhance teamperformance and innovation?

Possible additional topics for the minitrack include but are not limited to:

• AI-enabled code development tools
• AI-enabled team collaboration and communication tools
• New frontiers in agile or lean management – going beyond software development.
• Forecasting, planning, testing, measurement, and metrics
• Exploring the fit between agile (or lean) organizations and their environmental context
• Agile and lean requirements engineering, and risk management
• Agile in hybrid digital/physical contexts
• What cultures, team norms and leadership characteristics lead to sustained agility?
• Empirical studies of agile or lean organizations
• Impact of tool use on agile or lean management
• Education and training –new approaches to teaching and coaching agile
• Global software development and offshoring/multi-shoring
• Rapidly reconfigurable multi-sided platform ecosystems
• Project management methods, low code development

Minitrack Co-Chairs:

Jeffrey S. Saltz (Primary Contact)
Syracuse University
jsaltz@syr.edu

Edward G. Anderson
University of Texas at Austin
Edward.Anderson@mccombs.utexas.edu

Alex Sutherland
Scrum Master
alex.sutherland@scruminc.com

Viktoria Stray
University of Oslo
stray@ifi.uio.no

With the advancement of AI technology, AI algorithms start to match human performance for certain tasks (e.g. ChatGPT) and discover loopholes in systems that were not previously found.  AI in general and ML methods specifically are increasingly used with scientific data and applied with great promise to solve a broad variety of scientific problems.  With the increased use of AI comes an increase in inherent complexity.  Deep Learning (DL) models with billions of parameters, operating with very large data volumes on heterogeneous architectures, obscure their inner workings to human understanding.  Unlike traditional ML algorithms, such as rule-based decision trees or linear-regression models where the decision boundary is clear, interpreting a learned model is difficult.

The increased need for transparency is compounded by that of avoiding bias in predictions.  Numerous examples of bias have been discovered in image recognition, classification, and text generation. Formal explanations of how models achieve results, explicit representations of data, comprehensiveness and diversity of datasets used for training are crucial to foster trust in AI. Additionally, the accuracy of results obtained with AI is often the product of customization; experiments show that many are not reproducible at scale, even within expected error bounds. While reproducibility may not be needed for some uses of AI (e.g. when AI is used for the purpose of preliminary triage in drug discovery) in other uses, reproducible AI is paramount.    

Researchers need to understand how Artificial Intelligence (AI) and Machine Learning (ML) results are obtained in order to gain new insights and to establish confidence in the validity of these results. The promises of AI/ML will not be realized if scientists cannot trust the results, understand how they were obtained, gain transparency into what datasets, models and model parameters have been used or what features in the data lead to results. Like any good scientific results, AI/ML pipelines should be reproducible to the most possible extent. 

This minitrack will explore a number of themes related to explainable, reproducible, ethical, and trustworthy AI.  Papers will be published with conference proceedings. Topics of interest include but are not limited to the following:

 Computational and foundational methods for explaining AI models (XAI)

• Computational and foundational methods to ensure reproducibility of machine learning predictions

• Approaches discussing how Findable, Accessible, Interpretable, Re-usable (FAIR) principles can be applied to AI/ML including use of methods

• Mental models for interpreting AI results

• Computational and foundational methods and algorithms for detecting bias in AI/ML models

• Approaches, tools and best practices to perform sanity checks on data transformation pipelines

• Approaches, tools and best practices to keep track of experiments

• Use cases of explainability and reproducibility with AI models (XAI, RAI)

• Case studies of when AI models introduce bias in results

• Definitions and examples of trustworthy AI

Minitrack Co-Chairs: