Large Language Models (LLMs) aim to mimic a natural form of human conversation, likely contributing to an anthropomorphic perception of AI in contrast to conventional human-computer interfaces. Our study explores human-AI conversations and humans’ perception of their counterpart in a collaborative mystery solving task with Anthropic’s Claude 3.5 Sonnet v2 model. We collected self-report data on participants’ perception of the interaction, measured task performance, and analyzed conversational dynamics using LLM-based emotion coding. We found that humans’ perception of AI, ranging from that of a teammate or colleague to a tool, did not necessarily impact performance in mystery solving, but correlated with aspects of the interaction itself. When participants perceived the AI as a teammate or colleague, they felt a stronger sense of team cohesion and their conversations were more collaborative, with more positive emotions. These findings may help practitioners design human-AI interfaces that foster positive interactions without endangering performance.

In recent years, hybrid work has surged globally, making Hybrid Collaboration Spaces—the physical and virtual environments facilitating hybrid collaboration—increasingly relevant. However, this aspect is often not addressed in research and its scientific discourse is not always obviously apparent. In the existing body of literature, there are terminological ambiguities as well as temporal trends which make it hard to quickly identify relevant research. We thus present a systematic literature review on Hybrid Collaboration Spaces based on seven academic data sources (90 publications included, 996 screened). Our review encompasses several scientific fields and domains, contributing a detailed bibliometric report and a thorough thematic analysis highlighting challenges in hybrid collaboration and requirements for related spaces. We aim to 1) provide a concise overview of the research landscape, 2) advance the cross-domain dialog on hybrid collaboration, and 3) contribute to the disambiguation of terminology and establishment of a shared vocabulary.

Since 2020, worldwide COVID-19-related lockdowns have led to a rapid increase of remote collaboration, particularly in the domain of knowledge work. This has undoubtedly brought challenges (e.g., work-life boundary management, social isolation), but also opportunities. Practices that have proven successful (e.g., through increased task performance, efficiency or satisfaction) are worth retaining in future. In this qualitative empirical study, we analyzed four teams’ (14 participants in total) mandatory remote collaboration over a period of several days to several months during a nationally imposed lockdown. We report results derived from questionnaires, logbooks, group interviews, and meeting recordings. We identify possible factors influencing quality of task outcome as well as subjective aspects like satisfaction, motivation, and team atmosphere. As a basis for our conclusions, we provide a scheme for categorizing effects of remote collaboration based on an exhaustive literature review on pandemic-induced mandatory remote work and collaboration.

Following the vision of Industry 4.0, manufacturers started to break down products into subsystems, each of which is developed and supplied by different suppliers to tackle the increasing complexity of product development (e.g., in vehicle, machine or plant construction). This process needs to be supported by intuitive Computer-Aided Design (CAD) tools and virtual modeling and simulation environments to effectively meet the challenges of increasingly complex product development. Therefore, in this paper, we present some outcoming of a research project called DisMoSim (Distributed Modeling and Simulation), which proposes an intuitive virtual environment for modeling and simulation that embeds a Multi User Inverse Kinematic Solver (MUKS). The MUKS allows (i) individual subsystems to be tested intuitively, i.e., Rapid Testing (RT), and (ii) people not yet familiar with a subsystem to understand it easily and quickly through simple kinematic actions, i.e., Rapid Understanding (RU), without unnaturally interrupting the modeling process. Finally, (iii) the MUKS enables subsystems to be easily assembled into complete systems, i.e., Rapid Assembling (RA), by means of a so-called snapping-point mechanism which supports the complex process of aligning and assembling different subsystems. To evaluate these capabilities, we present the results of a user study conducted with 20 participants. Special attention has been given to the embedded MUKS in contrast to existing approaches known from common CAD and simulation software. Thus, we demonstrate the practicability, usability, and versatility of our approach.

In recent years, research on collaborative interaction has relied on manual coding of rich audio/video recordings. The fine-grained analysis of such material is extremely time-consuming and labor-intensive. This is not only difficult to scale, but, as a result, might also limit the quality and completeness of coding due to fatigue, inherent human biases, (accidental or intentional), and inter-rater inconsistencies. In this paper, we explore how recent advances in machine learning may reduce manual effort and loss of information while retaining the value of human intelligence in the coding process. We present ACACIA (AI Chain for Augmented Collaborative Interaction Analysis), an AI video data analysis application which combines a range of advances in machine perception of video material for the analysis of collaborative interaction. We evaluate ACACIA's abilities, show how far we can already get, and which challenges remain. Our contribution lies in establishing a combined machine and human analysis pipeline that may be generalized to different collaborative settings and guide future research.

Since the emergence of Bitcoin, blockchains (BCs) have been applied not only in the finance sector, but also in various other domains like health care, education or Industry 4.0 resulting in numerous different BC platforms and substantial research work. This plenty of efforts yielded to several valuable scientific surveys classifying and evaluating existing BC platforms. Although each of them puts forward a somewhat consolidated view on the field, it is still challenging to get rid of the "blockchain muddle" preventing even a common perception of the core functionality of BCs. Instead of providing yet another BC overview we conduct a meta survey of existing BC studies and report on lessons learned in this paper, being the basis for our vision towards a UML-based reference model considering both, structural and behavioral aspects of BCs and thereby identifying the nucleus of BCs.

Volunteering is an indispensable cornerstone of our society, covering nearly every part of our life, from social care to emergency management and education. This omnipresence of volunteering led to a plethora of volunteer management systems (VMS), mainly supporting NPOs in scheduling and allocating tasks to volunteers. In contrast to this NPO-centric approach of existing VMS, we focus on volunteers by investigating the following core question: "How can the engagement of volunteers be digitized and exploited in a life-long way". To provide a first step towards answering this question, the contribution of this paper is threefold. First, an in-depth study of related approaches is provided identifying shortcomings but also promising concepts. Second, challenges which have to be tackled to deal with the broad spectrum and peculiarities of volunteering are identified and a vision for a next-generation VMS called iVolunteer is pointed out. Third, promising technologies are identified and discussed in detail to lay the basis for the technical architecture of our envisioned iVolunteer-prototype.

Evolving metamodels are in the center of Model-Driven Engineering, necessitating the co-evolution of dependent artifacts like models and transformations. While model co-evolution has been extensively studied, transformation co-evolution has received less attention up to now. Current approaches for transformation co-evolution provide a fixed, restricted set of metamodel (MM) changes, only. Furthermore, composite changes are treated as monolithic units, which may lead to inconsistent co-evolution for overlapping atomic changes and prohibits extensibility. Finally, transformation co-evolution is considered in isolation, possibly inducing inconsistencies between model and transformation co-evolution. To overcome these limitations, we propose a complete set of atomic MM changes being able to describe arbitrary MM evolutions. Reusability and extensibility are supported by means of change composition, ensuring an intra-artifact consistent co-evolution. Furthermore, each change provides resolution actions for both, models and transformations, ensuring an inter-artifact consistent co-evolution. Based on our conceptual approach, a prototypical implementation is presented.

Model transformations play a vital role in Model-Driven Engineering. Due to their increasing complexity, proper means for ensuring their quality are needed. Although numerous approaches for testing of model transformations have been proposed, their focus is rather on formal verification than on execution-based testing. Additionally, existing approaches do not consider the actual transformation definition, rarely provide hints for debugging in case of an error and are specific to a certain transformation language. Therefore we propose TETRA Box as a generic framework for execution-based white-box testing of transformation languages. For this, we base on symbolic execution of model transformations. Additionally, by employing our Pattern-based Modeling Language for Model Transformations (PaMoMo), we are able to provide dedicated failure traces that are aligned to the actual transformation definition as a hint for debugging.

Model-Driven Engineering places models as first-class artifacts throughout the software lifecycle requiring the availability of proper model transformation languages. Although numerous languages are available, they lack convenient facilities for debugging and supporting understanding of the transformation logic. This is not least because the underlying transformation engines operate on a low level of abstraction, hiding the operational semantics of a high-level language. Consequently, low-level debugging information is available only, e.g., variable values. To tackle these limitations, we propose a DSL on top of Colored Petri Nets (CPNs) - called Transformation Nets (TNs) - for the execution and debugging of model transformations. By integrating all artifacts of a transformation, i.e., metamodel elements, model elements, and transformation logic, a runtime model for model transformations is provided, making the afore hidden operational semantics explicit. Based on this runtime model we present various means for debugging by means of an example showing how a QVT-Relations (QVT-R) specification may be debugged using TNs.

Model transformations play a key role in the vision of Model-Driven Engineering. Thereby, the resolution of structural heterogeneities between metamodels (MMs) represents the key challenge. For this task, current approaches require the definition of partly tricky, low-level recurring transformation logic but neglect to offer reusable components. Moreover, little attention has been paid to heterogeneities caused by the concept of inheritance, although extensively used in MMs. Therefore, we propose to specify model transformations in a plug & and play manner by a set of pre-defined mapping operators (MOps) representing a DSL to resolve structural heterogeneities. For coping with inheritance in MMs, we introduce an inheritance mechanism between MOps allowing to reuse parts of the mapping definitions. Moreover, dedicated MOps for resolving heterogeneities when one MM comprises inheritance hierarchies whereas the other one does not are presented, which are well-known problems in object-relational transformations and object-oriented refactorings.

In the Model-Driven Architecture (MDA) paradigm the Query/View/Transformation (QVT) standard plays a vital role for model transformations. Especially the high-level declarative QVT Relations language, however, has not yet gained widespread use in practice. This is not least due to missing tool support in general and inadequate debugging support in particular. Transformation engines interpreting QVT Relations operate on a low level of abstraction, hide the operational semantics of a transformation and scatter metamodels, models, QVT code, and trace information across different artifacts. We therefore propose a model-based debugger representing QVT Relations on bases of TROPIC, a model transformation language utilizing a variant of Colored Petri Nets (CPNs). As a prerequisite for convenient debugging, TROPIC provides a homogeneous view on all artifacts of a transformation on basis of a single formalism. Besides that, this formalism also provides a runtime model, thus making the afore hidden operational semantics of the transformation explicit. Using an explicit runtime model allows to employ model-based techniques for debugging, e.g., using the Object Constraint Language (OCL) for simply defining breakpoints and querying the execution state of a transformation.

Current research on context information focuses on the use of new kinds of sensors and on the aggregation and interpretation of sensor data. Having a closer look at applications supporting emergency scenarios with context information several problems arise. Emergency activities are complex coordination tasks which involve a multitude of different roles and resources. The definition of emergency scenarios and the mapping of relevant context information is a time-consuming and error-prone task.This work discusses how a model-based approach can support the definition of emergency scenarios at an abstract level. The abstract representation (PIM, Platform Independent Model) of an activity is then transformed into a platform-specific model (PSM), which includes a collection of context sensors. We show that the use of a model-based approach simplifies the definition of complex context-sensitive applications, and how this increases the flexibility to use different sensor platforms in different emergency scenarios.