Filtern
Erscheinungsjahr
- 2023 (8) (entfernen)
Dokumenttyp
- Dissertation (8) (entfernen)
Sprache
- Englisch (8) (entfernen)
Schlagworte
- Action Recognition (1)
- Action Segmentation (1)
- Anforderung (1)
- CLIL (1)
- Computer assisted communication (1)
- Curriculum (1)
- Design Science Research (1)
- Employee Behavior (1)
- Flesch-Reading-Ease Index (1)
- Fremdsprachendidaktik (1)
- Fächerkanon (1)
- Gamification (1)
- Imitation Learning (1)
- Leichte Sprache (1)
- Linguistic Requirements (1)
- Multimodal Action Recognition (1)
- One-Shot Action Recognition (1)
- Spiralcurriculum (1)
- Sustainability (1)
- empirische Untersuchung (1)
- lexical sophistication (1)
- natural language generation (1)
- plain language (1)
- scaffolded writing (1)
Leichte Sprache (LS) ist eine vereinfachte Varietät des Deutschen in der barrierefreie Texte für ein breites Spektrum von Menschen, einschließlich gering literalisierten Personen mit Lernschwierigkeiten, geistigen oder entwicklungsbedingten Behinderungen (IDD) und/oder komplexen Kommunikationsbedürfnissen (CCN), bereitgestellt werden. LS-Autor*innen sind i.d.R. der deutschen Standardsprache mächtig und gehören nicht der genannten Personengruppe an. Unser Ziel ist es, diese zu befähigen, selbst am schriftlichen Diskurs teilzunehmen. Hierfür bedarf es eines speziellen Schreibsystems, dessen linguistische Unterstützung und softwareergonomische Gestaltung den spezifischen Bedürfnissen der Zielgruppe gerecht wird. EasyTalk ist ein System basierend auf computerlinguistischer Verarbeitung natürlicher Sprache (NLP) für assistives Schreiben in einer erweiterten Variante von LS (ELS). Es stellt den Nutzenden ein personalisierbares Vokabular mit individualisierbaren Kommunikationssymbolen zur Verfügung und unterstützt sie entsprechend ihres persönlichen Fähigkeitslevels durch interaktive Benutzerführung beim Schreiben. Intuitive Formulierungen für linguistische Entscheidungen minimieren das erforderliche grammatikalische Wissen für die Erstellung korrekter und kohärenter komplexer Inhalte. Einfache Dialoge kommunizieren mit einem natürlichsprachlichen Paraphrasengenerator, der kontextsensitiv Vorschläge für Satzkomponenten und korrekt flektierte Wortformen bereitstellt. Außerdem regt EasyTalk die Nutzer*innen an, Textelemente hinzuzufügen, welche die Verständlichkeit des Textes für dessen Leserschaft fördern (z.B. Zeit- und Ortsangaben) und die Textkohärenz verbessern (z.B. explizite Diskurskonnektoren). Um das System auf die Bedürfnisse der Zielgruppe zuzuschneiden, folgte die Entwicklung von EasyTalk den Grundsätzen der menschzentrierten Gestaltung (UCD). Entsprechend wurde das System in iterativen Entwicklungszyklen ausgereift, kombiniert mit gezielten Evaluierungen bestimmter Aspekte durch Gruppen von Expert*innen aus den Bereichen CCN, LS und IT sowie L2-Lernende der deutschen Sprache. Eine Fallstudie, in welcher Mitglieder der Zielgruppe das freie Schreiben mit dem System testeten, bestätigte, dass Erwachsene mit geringen Lese-, Schreib- und Computerfähigkeiten mit IDD und/oder CCN mit EasyTalk eigene persönliche Texte in ELS verfassen können. Das positive Feedback aller Tests inspiriert Langzeitstudien mit EasyTalk und die Weiterentwicklung des prototypischen Systems, wie z.B. die Implementierung einer s.g. Schreibwerkstatt.
In the last years, the public interest in epidemiology and mathematical modeling of disease spread has increased - mainly caused by the COVID-19 pandemic, which has emphasized the urgent need for accurate and timely modelling of disease transmission. However, even prior to that, mathematical modelling has been used for describing the dynamics and spread of infectious diseases, which is vital for developing effective interventions and controls, e.g., for vaccination campaigns and social restrictions like lockdowns. The forecasts and evaluations provided by these models influence political actions and shape the measures implemented to contain the virus.
This research contributes to the understanding and control of disease spread, specifically for Dengue fever and COVID-19, making use of mathematical models and various data analysis techniques. The mathematical foundations of epidemiological modelling, as well as several concepts for spatio-temporal diffusion like ordinary differential equation (ODE) models, are presented, as well as an originally human-vector model for Dengue fever, and the standard (SEIR)-model (with the potential inclusion of an equation for deceased persons), which are suited for the description of COVID-19. Additionally, multi-compartment models, fractional diffusion models, partial differential equations (PDE) models, and integro-differential models are used to describe spatial propagation of the diseases.
We will make use of different optimization techniques to adapt the models to medical data and estimate the relevant parameters or finding optimal control techniques for containing diseases using both Metropolis and Lagrangian methods. Reasonable estimates for the unknown parameters are found, especially in initial stages of pandemics, when little to no information is available and the majority of the population has not got in contact with the disease. The longer a disease is present, the more complex the modelling gets and more things (vaccination, different types, etc.) appear and reduce the estimation and prediction quality of the mathematical models.
While it is possible to create highly complex models with numerous equations and parameters, such an approach presents several challenges, including difficulties in comparing and evaluating data, increased risk of overfitting, and reduced generalizability. Therefore, we will also consider criteria for model selection based on fit and complexity as well as the sensitivity of the model with respect to specific parameters. This also gives valuable information on which political interventions should be more emphasized for possible variations of parameter values.
Furthermore, the presented models, particularly the optimization using the Metropolis algorithm for parameter estimation, are compared with other established methods. The quality of model calculation, as well as computational effort and applicability, play a role in this comparison. Additionally, the spatial integro-differential model is compared with an established agent-based model. Since the macroscopic results align very well, the computationally faster integro-differential model can now be used as a proxy for the slower and non-traditionally optimizable agent-based model, e.g., in order to find an apt control strategy.
Durch die zunehmende Wichtigkeit und Dringlichkeit des Klimawandels sind Unternehmen aufgefordert, einen Beitrag zu nachhaltiger Entwicklung zu leisten, insbesondere durch die jüngeren Generationen. Bisherige Beiträge von Unternehmen werden jedoch als unzureichend kritisiert, was insbesondere am mangelnden Engagement der Mitarbeiterinnen und Mitarbeiter für Nachhaltigkeit in Unternehmen liegen könnte. In diesem Zusammenhang wurde in den letzten Jahren Gamification als ein vielversprechendes, innovatives Tool um nachhaltige Verhaltensweisen der Mitarbeiterinnen und Mitarbeiter am Arbeitsplatz zu motivieren, vorgeschlagen und zunehmend erforscht. Es gibt jedoch nur wenige Studien und anwendbare Gamification-Lösungen, die mehr als ein spezifisches Nachhaltigkeitsthema behandeln und somit eine ganzheitliche Perspektive auf nachhaltige Verhaltensweisen am Arbeitsplatz einnehmen. Darüber hinaus mangelt es bisheriger Forschung an einem umfassenden Verständnis dafür, wie verschiedene Gamification-Elemente spezifische psychologische Effekte hervorrufen, wie sich diese in Verhaltensänderungen manifestieren und wie diese wiederum kumulativ in messbaren Unternehmensergebnissen resultieren. Der Weg von Gamification als ”Input” zu unternehmerischer Nachhaltigkeit als ”Output” ist also bislang unerforscht.
Diese Dissertation schließt diese Lücke, indem eine ganzheitliche gamifizierte Intervention konzipiert, gestaltet und evaluiert wird, die Mitarbeiterinnen und Mitarbeiter bei verschiedenen nachhaltigen Verhaltensweisen in ihren täglichen Aktivitäten unterstützt. Das Projekt verwendet einen designwissenschaftlichen Forschungsansatz, der die Mitarbeiterinnen und Mitarbeiter eng in die schrittweise Entwicklung der Lösung einbezieht. Als Teil des iterativen Designprozesses werden in dieser Dissertation sechs Studien vorgestellt, um das theoretische Verständnis von Gamification für nachhaltige Verhaltensweisen von Mitarbeiterinnen und Mitarbeitern zu erweitern. Zunächst wird ein umfassender Überblick über die bestehende Forschung zu Gamification für nachhaltiges Mitarbeiterverhalten gegeben, wobei Gamification-Designs und Ergebnisse früherer Studien analysiert und eine Agenda für die weitere Forschung aufgezeigt werden (Studie 1). Danach werden theoretische Grundlagen der Forschung zu Gamification, Serious Games und Game-based Learning (Studie 2) und empirische Gestaltungsprinzipien
für Gamification und persuasive Systeme (Studie 3) als Basis für die erfolgreiche Gestaltung gamifizierter Anwendungen systematisch untersucht. Anschließend werden in empirischen Studien Motivationen der Mitarbeiterinnen und Mitarbeiter für nachhaltiges Verhalten erforscht und ihre Erwartungen an Gestaltungsmerkmale beleuchtet (Studie 4) sowie kontextuelle Herausforderungen und Gestaltungsdilemmata bei der Implementierung von Gamification in einem organisatorischen Kontext aufgezeigt (Studie 5). Schließlich wird in einer quantitativen Feldstudie (Studie 6) untersucht, wie verschiedene Gamification-Designs nachhaltiges Mitarbeiterverhalten und unternehmerische Nachhaltigkeitskennzahlen in Organisationen beeinflussen. Basierend auf den Ergebnissen wird in dieser Dissertation ein umfassendes Framework für Gamification für nachhaltiges Mitarbeiterverhalten präsentiert, welches Design-, individuelle Verhaltens- und Unternehmensperspektiven einbezieht. Schließlich werden darauf aufbauend praktische Empfehlungen für die Gestaltung von Gamification zur Förderung nachhaltigen Mitarbeiterverhaltens am Arbeitsplatz präsentiert.
On the recognition of human activities and the evaluation of its imitation by robotic systems
(2023)
This thesis addresses the problem of action recognition through the analysis of human motion and the benchmarking of its imitation by robotic systems.
For our action recognition related approaches, we focus on presenting approaches that generalize well across different sensor modalities. We transform multivariate signal streams from various sensors to a common image representation. The action recognition problem on sequential multivariate signal streams can then be reduced to an image classification task for which we utilize recent advances in machine learning. We demonstrate the broad applicability of our approaches formulated as a supervised classification task for action recognition, a semi-supervised classification task for one-shot action recognition, modality fusion and temporal action segmentation.
For action classification, we use an EfficientNet Convolutional Neural Network (CNN) model to classify the image representations of various data modalities. Further, we present approaches for filtering and the fusion of various modalities on a representation level. We extend the approach to be applicable for semi-supervised classification and train a metric-learning model that encodes action similarity. During training, the encoder optimizes the distances in embedding space for self-, positive- and negative-pair similarities. The resulting encoder allows estimating action similarity by calculating distances in embedding space. At training time, no action classes from the test set are used.
Graph Convolutional Network (GCN) generalized the concept of CNNs to non-Euclidean data structures and showed great success for action recognition directly operating on spatio-temporal sequences like skeleton sequences. GCNs have recently shown state-of-the-art performance for skeleton-based action recognition but are currently widely neglected as the foundation for the fusion of various sensor modalities. We propose incorporating additional modalities, like inertial measurements or RGB features, into a skeleton-graph, by proposing fusion on two different dimensionality levels. On a channel dimension, modalities are fused by introducing additional node attributes. On a spatial dimension, additional nodes are incorporated into the skeleton-graph.
Transformer models showed excellent performance in the analysis of sequential data. We formulate the temporal action segmentation task as an object detection task and use a detection transformer model on our proposed motion image representations. Experiments for our action recognition related approaches are executed on large-scale publicly available datasets. Our approaches for action recognition for various modalities, action recognition by fusion of various modalities, and one-shot action recognition demonstrate state-of-the-art results on some datasets.
Finally, we present a hybrid imitation learning benchmark. The benchmark consists of a dataset, metrics, and a simulator integration. The dataset contains RGB-D image sequences of humans performing movements and executing manipulation tasks, as well as the corresponding ground truth. The RGB-D camera is calibrated against a motion-capturing system, and the resulting sequences serve as input for imitation learning approaches. The resulting policy is then executed in the simulated environment on different robots. We propose two metrics to assess the quality of the imitation. The trajectory metric gives insights into how close the execution was to the demonstration. The effect metric describes how close the final state was reached according to the demonstration. The Simitate benchmark can improve the comparability of imitation learning approaches.
In a world where language defines the boundaries of one's understanding, the words of Austrian philosopher Ludwig Wittgenstein resonate profoundly. Wittgenstein's assertion that "Die Grenzen meine Sprache bedeuten die Grenzen meiner Welt" (Wittgenstein 2016: v. 5.6) underscores the vital role of language in shaping our perceptions. Today, in a globalized and interconnected society, fluency in foreign languages is indispensable for individual success. Education must break down these linguistic barriers, and one promising approach is the integration of foreign languages into content subjects.
Teaching content subjects in a foreign language, a practice known as Content Language Integrated Learning (CLIL), not only enhances language skills but also cultivates cognitive abilities and intercultural competence. This approach expands horizons and aligns with the core principles of European education (Leaton Gray, Scott & Mehisto 2018: 50). The Kultusministerkonferenz (KMK) recognizes the benefits of CLIL and encourages its implementation in German schools (cf. KMK 2013a).
With the rising popularity of CLIL, textbooks in foreign languages have become widely available, simplifying teaching. However, the appropriateness of the language used in these materials remains an unanswered question. If textbooks impose excessive linguistic demands, they may inadvertently limit students' development and contradict the goal of CLIL.
This thesis focuses on addressing this issue by systematically analyzing language requirements in CLIL teaching materials, emphasizing receptive and productive skills in various subjects based on the Common European Framework of Reference. The aim is to identify a sequence of subjects that facilitates students' language skill development throughout their school years. Such a sequence would enable teachers to harness the full potential of CLIL, fostering a bidirectional approach where content subjects facilitate language learning.
While research on CLIL is extensive, studies on language requirements for bilingual students are limited. This thesis seeks to bridge this gap by presenting findings for History, Geography, Biology, and Mathematics, allowing for a comprehensive understanding of language demands. This research endeavors to enrich the field of bilingual education and CLIL, ultimately benefiting the academic success of students in an interconnected world.
Die Aufmerksamkeit politischer Entscheidungsträger weltweit richtet sich in den letzten 10 Jahren verstärkt auf die Kreativwirtschaft als signifikanter Wachstums- und Beschäftigungsmotor in Städten. Die Literatur zeigt jedoch, dass Kreativschaffende zu den gefährdetsten Arbeitskräften in der heutigen Wirtschaft gehören. Aufgrund des enorm deregulierten und stark individualisierten Umfelds werden Misserfolg oder Erfolg eher individuellen Fähigkeiten und Engagement zugeschrieben und strukturelle oder kollektive Aspekte vernachlässigt. Diese Arbeit widmet sich zeitlichen, räumlichen und sozialen Aspekten digitaler behavioraler Daten, um zu zeigen, dass es tatsächlich strukturelle und historische Faktoren gibt, die sich auf die Karrieren von Individuen und Gruppen auswirken. Zu diesem Zweck bietet die Arbeit einen computergestützten, sozialwissenschaftlichen Forschungsrahmen, der das theoretische und empirisches Wissen aus jahrelanger Forschung zu Ungleichheit mit computergestützten Methoden zum Umgang mit komplexen und umfangreichen digitalen Daten verbindet. Die Arbeit beginnt mit der Darlegung einer neuartigen Methode zur Geschlechtererkennung, welche sich Image Search und Gesichtserkennungsmethoden bedient. Die Analyse der kollaborativen Verhaltensweisen sowie der Zitationsnetzwerke männlicher und weiblicher Computerwissenschaftler*innen verdeutlicht einige der historischen Bias und Nachteile, welchen Frauen in ihren wissenschaftlichen Karrieren begegnen. Zur weiterfuhrenden Elaboration der zeitlichen Aspekte von Ungleichheit, wird der Anteil vertikaler und horizontaler Ungleichheit in unterschiedlichen Kohorten von Wissenschaftler*innen untersucht, die ihre Karriere zu unterschiedlichen Zeitpunkten begonnen haben. Im Weiteren werden einige der zugrunde liegenden Mechanismen und Prozesse von Ungleichheit in kreativen Berufen analysiert, wie der Matthew-Effekt und das Hipster-Paradoxon. Schließlich zeigt diese Arbeit auf, dass Online-Plattformen wie Wikipedia bestehenden Bias reflektieren sowie verstärken können.
The diversity within amphibian communities in cultivated areas in Rwanda and within two selected, taxonomically challenging groups, the genera Ptychadena and Hyperolius, were investigated in this thesis. The amphibian community of an agricultural wetland near Butare in southern Rwanda comprised 15 anuran species. Rarefaction and jackknife analyses corroborated that the complete current species richness of the assemblage had been recorded, and the results of acoustic niche analysis suggested species saturation of the community. Surveys at many other Rwandan localities showed that the species recorded in Butare are widespread in cultivated and pristine wetlands. The species were readily distinguishable using morphological, bioacoustic, and molecular (DNA barcoding) features, but only eight of the 15 species could be assigned unambiguously to nominal species. The remaining represented undescribed or currently unrecognized taxa, including three species of Hyperolius, two Phrynobatrachus species, one Ptychadena species, and one species of Amietia. The diversity of the Ridged Frogs in Rwanda was investigated in two studies (Chapters III and IV). Three species of Ptychadena were recorded in wetlands in the catchment of the Nile. They can be distinguished by morphological characters (morphometrics and qualitative features) as well as by their advertisement calls and genetics. The Rwandan species of the P. mascareniensis group was shown to differ from the topotypic population as well as from other genetic lineages in sub-Saharan Africa and an old available name, P. nilotica, was resurrected from synonymy for this lineage. Two further Ptychadena species were identified among voucher specimens from Rwanda deposited in the collection of the RMCA, P. chrysogaster and P. uzungwensis. Morphologically they can be unambiguously distinguished from each other and the three other Rwandan species. A key based on qualitative morphological characters was developed, which allows unequivocal identification of specimens of all species that have been recorded from Rwanda. DNA was isolated from a Rwandan voucher specimen of P. chrysogaster, and the genetic analysis corroborated the species" distinct status.
A species of Hyperolius collected in the Nyungwe National Park was compared to all other Rwandan species of the genus and to morphologically or genetically similar species from neighbouring countries. Its distinct taxonomic status was justified by morphological, bioacoustic, and molecular evidence and it was described as a new species, H. jackie. A species of the H. nasutus group collected at agricultural sites in Rwanda was described as a new species in the course of a revision of the species of the Hyperolius nasutus group. The group was shown to consist of 15 distinct species which can be distinguished from each other genetically, bioacoustically, and morphologically.
The aerial performance, i.e. parachuting, of the Disc-fingered Reed Frog, Hyperolius discodactylus, was described. It represents a novel observation of a behaviour that has been known from a number of Southeast Asian and Neotropical frog species. Parachuting frogs, including H. discodactylus, exhibit certain morphological characteristics and, while airborne, assume a distinct posture which is best-suited for maneuvering in the air. Another study on the species addressed the validity of the taxon H. alticola which had been considered either a synonym of H. discodactylus or a distinct species. Type material of both taxa was re-examined and the status of H. alticola reassessed using morphological data from historic and new collections, call recordings, and molecular data from animals collected on recent expeditions. A northern and a southern genetic clade were identified, a divide that is weakly supported by diverging morphology of the vouchers from the respective localities. No distinction in advertisement call features could be recovered to support this split and both genetic and morphological differences between the two geographic clades are marginal and not always congruent and more likely reflect population-level variation. Therefore it was concluded that H. alticola is not a valid taxon and should be treated as a synonym of H. discodactylus.
The trends of industry 4.0 and the further enhancements toward an ever changing factory lead to more mobility and flexibility on the factory floor. With that higher need of mobility and flexibility the requirements on wireless communication rise. A key requirement in that setting is the demand for wireless Ultra-Reliability and Low Latency Communication (URLLC). Example use cases therefore are cooperative Automated Guided Vehicles (AGVs) and mobile robotics in general. Working along that setting this thesis provides insights regarding the whole network stack. Thereby, the focus is always on industrial applications. Starting on the physical layer, extensive measurements from 2 GHz to 6 GHz on the factory floor are performed. The raw data is published and analyzed. Based on that data an improved Saleh-Valenzuela (SV) model is provided. As ad-hoc networks are highly depended onnode mobility, the mobility of AGVs is modeled. Additionally, Nodal Encounter Patterns (NEPs) are recorded and analyzed. A method to record NEP is illustrated. The performance by means of latency and reliability are key parameters from an application perspective. Thus, measurements of those two parameters in factory environments are performed using Wireless Local Area Network (WLAN) (IEEE 802.11n), private Long Term Evolution (pLTE) and 5G. This showed auto-correlated latency values. Hence, a method to construct confidence intervals based on auto-correlated data containing rare events is developed. Subsequently, four performance improvements for wireless networks on the factory floor are proposed. Of those optimization three cover ad-hoc networks, two deal with safety relevant communication, one orchestrates the usage of two orthogonal networks and lastly one optimizes the usage of information within cellular networks.
Finally, this thesis is concluded by an outlook toward open research questions. This includes open questions remaining in the context of industry 4.0 and further the ones around 6G. Along the research topics of 6G the two most relevant topics concern the ideas of a network of networks and overcoming best-effort IP.