Data scarcity is one of the main problems when it comes to real-world applications of transformer-based models. This is especially evident for task-oriented dialogue (TOD) systems, which require specialized datasets, that are usually not readily available. This can hinder companies from adding TOD systems to their services. This study therefore investigates a novel approach to sourcing annotated dialogues from existing German monologue material. Focusing on a real-world example, we investigate whether these monologues can be transformed into dialogue formats suitable for training TOD systems. We show the approach with the concrete example of a company specializing in travel bookings via e-mail. We fine-tune state-of-the-art Large Language Models for the task of rewriting e-mails as dialogues and annotating them. To ensure the quality and validity of the generated data, we employ crowd workers to evaluate the dialogues across multiple criteria and to provide gold-standard annotations for the test dataset. We further evaluate the usefulness of the dialogues for training TOD systems. Our evaluation shows that the dialogues and annotations are of high quality and can serve as a valuable starting point for training TOD systems. Finally, we make the annotated dataset publicly available to foster future research.

While it is often assumed that searching for information to evaluate misinformation will help identify false claims, recent work suggests that search behaviours can instead reinforce belief in misleading news, particularly when users generate queries using vocabulary from the source articles. Our research explores how different query generation strategies affect news verification and whether the way people search influences the accuracy of their information evaluation. A mixed-methods approach was used, consisting of three parts: (1) an analysis of existing data to understand how search behaviour influences trust in fake news, (2) a simulation of query generation strategies using a Large Language Model (LLM) to assess the impact of different query formulations on search result quality, and (3) a user study to examine how 'Boost' interventions in interface design can guide users to adopt more effective query strategies. The results show that search behaviour significantly affects trust in news, with successful searches involving multiple queries and yielding higher-quality results. Queries inspired by different parts of a news article produced search results of varying quality, and weak initial queries improved when reformulated using full SERP information. Although 'Boost' interventions had limited impact, the study suggests that interface design encouraging users to thoroughly review search results can enhance query formulation. This study highlights the importance of query strategies in evaluating news and proposes that interface design can play a key role in promoting more effective search practices, serving as one component of a broader set of interventions to combat misinformation.

Large Language Models (LLMs) have demonstrated remarkable capabilities in conversational tasks. Embodying an LLM as a virtual human allows users to engage in face-to-face social interactions in Virtual Reality. However, the influence of person- and task-related factors in social interactions with LLM-controlled agents remains unclear. In this study, forty-six participants interacted with a virtual agent whose persona was manipulated as extravert or introvert in three different conversational tasks (small talk, knowledge test, convincing). Social-evaluation, emotional experience, and realism were assessed using ratings. Interactive engagement was measured by quantifying participants' words and conversational turns. Finally, we measured participants' willingness to ask the agent for help during the knowledge test. Our findings show that the extraverted agent was more positively evaluated, elicited a more pleasant experience and greater engagement, and was assessed as more realistic compared to the introverted agent. Whereas persona did not affect the tendency to ask for help, participants were generally more confident in the answer when they had help of the LLM. Variation of personality traits of LLM-controlled embodied virtual agents, therefore, affects social-emotional processing and behavior in virtual interactions. Embodied virtual agents allow the presentation of naturalistic social encounters in a virtual environment.

Large Language Models (LLMs) have recently been studied within the context of Task-Oriented Dialogues (TOD). However, previous research is inconclusive on their effectiveness, with some studies claiming that LLMs are unable to perform the TOD task and others making sophisticated additions to their setup and coming to opposite conclusions. In this work, we take a detailed look at previous results that state LLMs perform insufficiently as a TOD system. As a result, we propose an updated, stronger baseline for multiple out-of-the-box LLM performances as TOD systems. We introduce a Self-Checking mechanism as a simple, yet effective, component to drastically improve their performance. Our results show that newer, pre-trained LLMs can, in fact, perform as TOD systems out-of-the-box, challenging the previous understanding. We show that LLMs can even perform competitively to fine-tuned models in certain metrics. Based on this, we propose directions for future research. Our code is published on Github.

Large-scale Wizard-Of-Oz dialogue datasets have enabled the training of deep learning-based dialogue systems. While they are successful as benchmark datasets, they lack certain types of utterances, which would make them more realistic. In this work, we investigate the creation of synthetic communication errors in an automatic pipeline. Based on linguistic theory, we propose and follow a simple error taxonomy. We focus on three types of miscommunications that could happen in real-world dialogues but are underrepresented in the benchmark dataset: misunderstandings, non-understandings and vaguely related questions. Our two-step approach uses a state-of-the-art Large Language Model (LLM) to first create the error and secondly the repairing utterance. We perform Language Model-based evaluation to ensure the quality of the generated utterances. We apply the method to the MultiWOZ dataset and evaluate it both qualitatively and empirically as well as with human judges. Our results indicate that current LLMs can aid in adding post-hoc miscommunications to benchmark datasets as a form of data augmentation. We publish the resulting dataset, in which nearly 1900 dialogues have been modified, as CoPrUS-MultiWOZ to facilitate future work on dialogue systems.

Indoor navigation systems for pedestrians require floor plans of buildings. Acquiring them is costly and time-consuming, as the standard procedure is to digitise 2D blueprints and augment them with information required for navigating users. In this demo, as an alternative, we use the Holo Lens 2 to construct floor plans on site. We illustrate how to construct plans that are as accurate as digitised blueprints, but much more efficient to acquire and easier to augment with information about the physical environment that is impossible to obtain from blueprints.

Even though assistance systems offer more potential due to the increasing maturity of the inherent technologies, Automatic Speech Recognition faces distinctive challenges in the industrial context. Speech recognition enables immersive assistance systems to handle inputs and commands hands-free during two-handed operative jobs. The results of the conducted study (with n = 22 participants) based on the counterbalanced within-subject design demonstrated the performance (word error rate and information transfer rate) of the HMD HoloLens 2 as a function of the sound pressure level of industrial noise. The negative influence of industrial noise was higher on the word error rate of dictation than on the information transfer rate of the speech command. Contrary to expectations, no statistically significant difference in performance was found between the stationary and non-stationary noise. Furthermore, this study confirmed the hypothesis that user acceptance was negatively influenced by erroneous speech interactions. Furthermore, the erroneous speech interaction had no statistically significant influence on the workload or physiological parameters (skin conductance level and heart rate). It can be summarized that Automatic Speech Recognition is not yet a capable interaction paradigm in an industrial context.

Emerging assistance systems are designed to enable operators to perform tasks better, faster, and with a lower workload. However, in line with the productivity paradox, the full potential of automation and digitalisation is not being realised. One reason for this is insufficient training. In this study, the statistically significant differences among three different training scenarios on performance, acceptance, workload, and technostress during the execution of immersive measurement tasks are demonstrated. A between-subjects design was applied and analysed using ANOVAs involving 52 participants (with a statistical overall power of 0.92). The ANOVAs were related to three levels of the independent variable: quality training, manipulated as minimal, personal, and optimised training. The results show that the quality of training significantly influences immersive assistance systems. Hence, this article deduces tangible design guidelines for training, with consideration of the system-level hardware, operational system, and immersive application. Surprisingly, an appropriate mix of training approaches, rather than detailed, personalised training, appears to be more effective than e-learning or ‘getting started’ tools for immersive systems. In contrast to most studies in the related work, our article is not about learning with AR applications but about training scenarios for the use of immersive systems.

Recently, the research into language models fine-tuned to follow prompts has made notable advances. These are commonly used in the form of chatbots. One special case of chatbots is that of Task-Oriented Dialogue (TOD) systems that aim to help the user achieve specific tasks using external services. High quality training data for these systems is costly to come by. We thus evaluate if the new prompt-following models can generate annotated synthetic dialogues and if these can be used to train a TOD system. To this end we generate data based on descriptions of a dialogues goal. We train a state-of-the-art TOD system to compare it in a low resource setting with and without synthetic dialogues. The evaluation shows that using prompt-following language models to generate synthetic dialogues could help training better TOD systems.

Modern dialog systems rely on Deep Learning to train transformer-based model architectures. These notoriously rely on large amounts of training data. However, the collection of conversational data is often a tedious and costly process. This is especially true for Task-Oriented Dialogs, where the system ought to help the user achieve specific tasks, such as making reservations. We investigate a controlled strategy for dialog synthesis. Our method generates utterances based on dialog annotations in a sequence-to-sequence manner. Besides exploring the viability of the approach itself, we also explore the effect of constrained beam search on the generation capabilities. Moreover, we analyze the effectiveness of the proposed method as a data augmentation by studying the impact the synthetic dialogs have on training dialog systems. We perform the experiments in multiple settings, simulating various amounts of ground-truth data. Our work shows that a controlled generation approach is a viable method to synthesize Task-Oriented Dialogs, that can in turn be used to train dialog systems. We were able to improve this process by utilizing constrained beam search.

In this report, we present the project URWalking conducted at the University of Regensburg. We describe its major outcomes: Firstly, an indoor navigation system for pedestrians as a web application and as an Android app with position tracking of users in indoor and outdoor environments. Our implementation showcases that a variant of the $$A^*$$ -algorithm by Ullmann (tengetriebene optimierung präferenzadaptiver fußwegrouten durch gebäudekomplexe https://epub.uni-regensburg.de/43697/ , 2020) can handle the routing problem in large, levelled indoor environments efficiently. Secondly, the apps have been used in several studies for a deeper understanding of human wayfinding. We collected eye tracking and synchronized video data, think aloud protocols, and log data of users interacting with the apps. We applied state-of-the-art deep learning models for gaze tracking and automatic classification of landmarks. Our results indicate that even the most recent version of the YOLO image classifier by Redmon and Farhadi (olov3: An incremental improvement. arXiv, 2018) needs finetuning to recognize everyday objects in indoor environments. Furthermore, we provide empirical evidence that appropriate machine learning models are helpful to bridge behavioural data from users during wayfinding and conceptual models for the salience of objects and landmarks. However, simplistic models are insufficient to reasonably explain wayfinding behaviour in real time—an open issue in GeoAI. We conclude that the GeoAI community should collect more naturalistic log data of wayfinding activities in order to build efficient machine learning models capable of predicting user reactions to routing instructions and of explaining how humans integrate stimuli from the environment as essential information into routing instructions while solving wayfinding tasks. Such models form the basis for real-time wayfinding assistance.

As Kant shows in A 71-76 of the First Critique, his table of the twelve "logical functions of understanding" (in A 70) is an indispensable extension of a table of four well-known logical functions that we find in a section of the Logic that was "already finished" in Aristotle's times: The Square of Oppositions. The undisputed completeness of this special table thus warrants the completeness of Kant's general table as well. Any further philosophical proof of completeness for Kant's table of judgements as a whole is therefore not necessary at all. And due to the contingency of "kind and number" of human forms of intuition and functions of judgment, such a 'proof' would not even be possible according to Kant - and thus it is not a subject (or even a part) of his Transcendental Deduction of the Categories. A concluding evaluation of Kant's own statements about the proof-structure of the B-Deduction as a whole supports this claim.

Abstract As Kant shows in A 71–76 of the First Critique, his table of the twelve “logical functions of understanding” (in A 70) is an indispensable extension of a table of four well-known logical functions that we find in a section of the Logic that was “already finished” in Aristotle’s times: The Square of Oppositions. The undisputed completeness of this special table thus warrants the completeness of Kant’s general table as well. Any further philosophical proof of completeness for Kant’s table of judgements as a whole is therefore not necessary at all. And due to the contingency of “kind and number” of human forms of intuition and functions of judgment, such a ‘proof’ would not even be possible according to Kant – and thus it is not a subject (or even a part) of his Transcendental Deduction of the Categories. A concluding evaluation of Kant’s own statements about the proof-structure of the B-Deduction as a whole supports this claim.

Quality of data and the assessment of data value have become essential requirements to compete in the knowledge economy. In this work, information needs and data requirements of three different user groups from the semiconductor manufacturing area were determined empirically, based on which a new Data Value Assessment (DVA) process could be designed. Three data collection phases were carried out, in which a total of 24 employees of ams-OSRAM International GmbH took part. First, variables that are relevant for the evaluation of data were identified by focus groups. These results made it possible to develop a basic DVA process. In the second step, the process was tested by means of a survey. Based on 14 data objects from a selected case project, an initial DVA was carried out by the user groups of managers, operational employees and the Data Governance Office. The survey also covered questions about the information behavior of the user groups. With the results of the questionnaire, it was possible to improve the DVA process and to identify roles during the data evaluation. The DVA process has been assessed against practical applications based on feedback from key stakeholders and the findings provide a clear roadmap for future directions.

Quality of data and the assessment of data value have become essential requirements to compete in the knowledge economy. In this work, information needs and data requirements of three different user groups from the semiconductor manufacturing area were determined empirically, based on which a new Data Value Assessment (DVA) process could be designed. Three data collection phases were carried out, in which a total of 24 employees of ams-OSRAM International GmbH took part. First, variables that are relevant for the evaluation of data were identified by focus groups. These results made it possible to develop a basic DVA process. In the second step, the process was tested by means of a survey. Based on 14 data objects from a selected case project, an initial DVA was carried out by the user groups of managers, operational employees and the Data Governance Office. The survey also covered questions about the information behavior of the user groups. With the results of the questionnaire, it was possible to improve the DVA process and to identify roles during the data evaluation. The DVA process has been assessed against practical applications based on feedback from key stakeholders and the findings provide a clear roadmap for future directions.