Refine
Year of publication
- 2023 (92) (remove)
Institute
- Fachbereich Medizintechnik und Technomathematik (22)
- Fachbereich Elektrotechnik und Informationstechnik (20)
- ECSM European Center for Sustainable Mobility (16)
- Fachbereich Luft- und Raumfahrttechnik (16)
- Fachbereich Chemie und Biotechnologie (10)
- Fachbereich Energietechnik (10)
- Fachbereich Wirtschaftswissenschaften (7)
- IfB - Institut für Bioengineering (7)
- Fachbereich Maschinenbau und Mechatronik (6)
- MASKOR Institut für Mobile Autonome Systeme und Kognitive Robotik (6)
Has Fulltext
- no (92) (remove)
Language
- English (92) (remove)
Document Type
- Article (43)
- Conference Proceeding (35)
- Part of a Book (6)
- Habilitation (2)
- Preprint (2)
- Book (1)
- Conference: Meeting Abstract (1)
- Contribution to a Periodical (1)
- Talk (1)
Keywords
- Natural language processing (3)
- Associated liquids (2)
- Diversity Management (2)
- Engineering Habitus (2)
- Future Skills (2)
- Information extraction (2)
- Interdisciplinarity (2)
- Organizational Culture (2)
- Power plants (2)
- Sustainability (2)
Preprint: Studies on the enzymatic reduction of levulinic acid using Chiralidon-R and Chiralidon-S
(2023)
The enzymatic reduction of levulinic acid by the chiral catalysts Chiralidon-R and Chiralidon-S which are commercially available superabsorbed alcohol dehydrogenases is described. The Chiralidon®-R/S reduces the levulinic acid to the (R,S)-4-hydroxy valeric acid and the (R)- or (S)- gamma-valerolactone.
In recent years, the development of large pretrained language models, such as BERT and GPT, significantly improved information extraction systems on various tasks, including relation classification. State-of-the-art systems are highly accurate on scientific benchmarks. A lack of explainability is currently a complicating factor in many real-world applications. Comprehensible systems are necessary to prevent biased, counterintuitive, or harmful decisions.
We introduce semantic extents, a concept to analyze decision patterns for the relation classification task. Semantic extents are the most influential parts of texts concerning classification decisions. Our definition allows similar procedures to determine semantic extents for humans and models. We provide an annotation tool and a software framework to determine semantic extents for humans and models conveniently and reproducibly. Comparing both reveals that models tend to learn shortcut patterns from data. These patterns are hard to detect with current interpretability methods, such as input reductions. Our approach can help detect and eliminate spurious decision patterns during model development. Semantic extents can increase the reliability and security of natural language processing systems. Semantic extents are an essential step in enabling applications in critical areas like healthcare or finance. Moreover, our work opens new research directions for developing methods to explain deep learning models.
Extracting workflow nets from textual descriptions can be used to simplify guidelines or formalize textual descriptions of formal processes like business processes and algorithms. The task of manually extracting processes, however, requires domain expertise and effort. While automatic process model extraction is desirable, annotating texts with formalized process models is expensive. Therefore, there are only a few machine-learning-based extraction approaches. Rule-based approaches, in turn, require domain specificity to work well and can rarely distinguish relevant and irrelevant information in textual descriptions. In this paper, we present GUIDO, a hybrid approach to the process model extraction task that first, classifies sentences regarding their relevance to the process model, using a BERT-based sentence classifier, and second, extracts a process model from the sentences classified as relevant, using dependency parsing. The presented approach achieves significantly better resul ts than a pure rule-based approach. GUIDO achieves an average behavioral similarity score of 0.93. Still, in comparison to purely machine-learning-based approaches, the annotation costs stay low.
The work in modern open-pit and underground mines requires the transportation of large amounts of resources between fixed points. The navigation to these fixed points is a repetitive task that can be automated. The challenge in automating the navigation of vehicles commonly used in mines is the systemic properties of such vehicles. Many mining vehicles, such as the one we have used in the research for this paper, use steering systems with an articulated joint bending the vehicle’s drive axis to change its course and a hydraulic drive system to actuate axial drive components or the movements of tippers if available. To address the difficulties of controlling such a vehicle, we present a model-predictive approach for controlling the vehicle. While the control optimisation based on a parallel error minimisation of the predicted state has already been established in the past, we provide insight into the design and implementation of an MPC for an articulated mining vehicle and show the results of real-world experiments in an open-pit mine environment.
Supervised machine learning and deep learning require a large amount of labeled data, which data scientists obtain in a manual, and time-consuming annotation process. To mitigate this challenge, Active Learning (AL) proposes promising data points to annotators they annotate next instead of a subsequent or random sample. This method is supposed to save annotation effort while maintaining model performance.
However, practitioners face many AL strategies for different tasks and need an empirical basis to choose between them. Surveys categorize AL strategies into taxonomies without performance indications. Presentations of novel AL strategies compare the performance to a small subset of strategies. Our contribution addresses the empirical basis by introducing a reproducible active learning evaluation (ALE) framework for the comparative evaluation of AL strategies in NLP.
The framework allows the implementation of AL strategies with low effort and a fair data-driven comparison through defining and tracking experiment parameters (e.g., initial dataset size, number of data points per query step, and the budget). ALE helps practitioners to make more informed decisions, and researchers can focus on developing new, effective AL strategies and deriving best practices for specific use cases. With best practices, practitioners can lower their annotation costs. We present a case study to illustrate how to use the framework.
Messenger apps like WhatsApp and Telegram are frequently used for everyday communication, but they can also be utilized as a platform for illegal activity. Telegram allows public groups with up to 200.000 participants. Criminals use these public groups for trading illegal commodities and services, which becomes a concern for law enforcement agencies, who manually monitor suspicious activity in these chat rooms. This research demonstrates how natural language processing (NLP) can assist in analyzing these chat rooms, providing an explorative overview of the domain and facilitating purposeful analyses of user behavior. We provide a publicly available corpus of annotated text messages with entities and relations from four self-proclaimed black market chat rooms. Our pipeline approach aggregates the extracted product attributes from user messages to profiles and uses these with their sold products as features for clustering. The extracted structured information is the foundation for further data exploration, such as identifying the top vendors or fine-granular price analyses. Our evaluation shows that pretrained word vectors perform better for unsupervised clustering than state-of-the-art transformer models, while the latter is still superior for sequence labeling.
To fulfil the CO2 emission reduction targets of the European Union (EU), heavy-duty (HD) trucks need to operate 15% more efficiently by 2025 and 30% by 2030. Their electrification is necessary as conventional HD trucks are already optimized for the long-haul application. The resulting hybrid electric vehicle (HEV) truck gains most of the fuel saving potential by the recuperation of potential energy and its consecutive utilization. The key to utilizing the full potential of HEV-HD trucks is to maximize the amount of recuperated energy and ensure its intelligent usage while keeping the operating point of the internal combustion engine as efficient as possible. To achieve this goal, an intelligent energy management strategy (EMS) based on ECMS is developed for a parallel HEV-HD truck which uses predictive discharge of the battery and adaptive operating strategy regarding the height profile and the vehicle mass. The presented EMS can reproduce the global optimal operating strategy over long phases and lead to a fuel saving potential of up to 2% compared with a heuristic strategy. Furthermore, the fuel saving potential is correlated with the investigated boundary conditions to deepen the understanding of the impact of intelligent EMS for HEV-HD trucks.
Using scenarios is vital in identifying and specifying measures for successfully transforming the energy system. Such transformations can be particularly challenging and require the support of a broader set of stakeholders. Otherwise, there will be opposition in the form of reluctance to adopt the necessary technologies. Usually, processes for considering stakeholders' perspectives are very time-consuming and costly. In particular, there are uncertainties about how to deal with modifications in the scenarios. In principle, new consulting processes will be required. In our study, we show how multi-criteria decision analysis can be used to analyze stakeholders' attitudes toward transition paths. Since stakeholders differ regarding their preferences and time horizons, we employ a multi-criteria decision analysis approach to identify which stakeholders will support or oppose a transition path. We provide a flexible template for analyzing stakeholder preferences toward transition paths. This flexibility comes from the fact that our multi-criteria decision aid-based approach does not involve intensive empirical work with stakeholders. Instead, it involves subjecting assumptions to robustness analysis, which can help identify options to influence stakeholders' attitudes toward transitions.
This study evaluates neuromechanical control and muscle-tendon interaction during energy storage and dissipation tasks in hypergravity. During parabolic flights, while 17 subjects performed drop jumps (DJs) and drop landings (DLs), electromyography (EMG) of the lower limb muscles was combined with in vivo fascicle dynamics of the gastrocnemius medialis, two-dimensional (2D) kinematics, and kinetics to measure and analyze changes in energy management. Comparisons were made between movement modalities executed in hypergravity (1.8 G) and gravity on ground (1 G). In 1.8 G, ankle dorsiflexion, knee joint flexion, and vertical center of mass (COM) displacement are lower in DJs than in DLs; within each movement modality, joint flexion amplitudes and COM displacement demonstrate higher values in 1.8 G than in 1 G. Concomitantly, negative peak ankle joint power, vertical ground reaction forces, and leg stiffness are similar between both movement modalities (1.8 G). In DJs, EMG activity in 1.8 G is lower during the COM deceleration phase than in 1 G, thus impairing quasi-isometric fascicle behavior. In DLs, EMG activity before and during the COM deceleration phase is higher, and fascicles are stretched less in 1.8 G than in 1 G. Compared with the situation in 1 G, highly task-specific neuromuscular activity is diminished in 1.8 G, resulting in fascicle lengthening in both movement modalities. Specifically, in DJs, a high magnitude of neuromuscular activity is impaired, resulting in altered energy storage. In contrast, in DLs, linear stiffening of the system due to higher neuromuscular activity combined with lower fascicle stretch enhances the buffering function of the tendon, and thus the capacity to safely dissipate energy.
The popularity of social media and particularly Instagram grows steadily. People use the different platforms to share pictures as well as videos and to communicate with friends. The potential of social media platforms is also being used for marketing purposes and for selling products. While for Facebook and other online social media platforms the purchase decision factors are investigated several times, Instagram stores remain mainly unattended so far. The present research work closes this gap and sheds light into decisive factors for purchasing products offered in Instagram stores. A theoretical research model, which contains selected constructs that are assumed to have a significant influence on Instagram user´s purchase intention, is developed. The hypotheses are evaluated by applying structural equation modelling on survey data containing 127 relevant participants. The results of the study reveal that ‘trust’, ‘personal recommendation’, and ‘usability’ significantly influences user’s buying intention in Instagram stores.