Refine
Year of publication
- 2024 (25)
- 2023 (101)
- 2022 (132)
- 2021 (145)
- 2020 (157)
- 2019 (192)
- 2018 (168)
- 2017 (154)
- 2016 (154)
- 2015 (176)
- 2014 (166)
- 2013 (171)
- 2012 (154)
- 2011 (184)
- 2010 (179)
- 2009 (185)
- 2008 (155)
- 2007 (149)
- 2006 (160)
- 2005 (130)
- 2004 (161)
- 2003 (106)
- 2002 (130)
- 2001 (106)
- 2000 (108)
- 1999 (109)
- 1998 (99)
- 1997 (99)
- 1996 (81)
- 1995 (78)
- 1994 (86)
- 1993 (59)
- 1992 (54)
- 1991 (29)
- 1990 (39)
- 1989 (45)
- 1988 (57)
- 1987 (32)
- 1986 (19)
- 1985 (34)
- 1984 (22)
- 1983 (20)
- 1982 (29)
- 1981 (20)
- 1980 (36)
- 1979 (24)
- 1978 (34)
- 1977 (14)
- 1976 (13)
- 1975 (12)
- 1974 (3)
- 1973 (2)
- 1972 (2)
- 1971 (1)
- 1968 (1)
Document Type
- Article (3226)
- Conference Proceeding (1146)
- Part of a Book (184)
- Book (144)
- Doctoral Thesis (30)
- Patent (25)
- Other (9)
- Report (9)
- Working Paper (6)
- Lecture (5)
- Poster (4)
- Preprint (4)
- Talk (4)
- Master's Thesis (2)
- Bachelor Thesis (1)
- Contribution to a Periodical (1)
- Habilitation (1)
Language
- English (4801) (remove)
Keywords
- Biosensor (25)
- Finite-Elemente-Methode (12)
- Einspielen <Werkstoff> (10)
- CAD (8)
- civil engineering (8)
- Bauingenieurwesen (7)
- Blitzschutz (6)
- FEM (6)
- Gamification (6)
- Limit analysis (6)
- Shakedown analysis (6)
- avalanche (6)
- shakedown analysis (6)
- Clusterion (5)
- Earthquake (5)
- Enterprise Architecture (5)
- MINLP (5)
- solar sail (5)
- Air purification (4)
- Diversity Management (4)
Institute
- Fachbereich Medizintechnik und Technomathematik (1668)
- Fachbereich Elektrotechnik und Informationstechnik (693)
- IfB - Institut für Bioengineering (620)
- Fachbereich Energietechnik (579)
- INB - Institut für Nano- und Biotechnologien (555)
- Fachbereich Chemie und Biotechnologie (534)
- Fachbereich Luft- und Raumfahrttechnik (477)
- Fachbereich Maschinenbau und Mechatronik (278)
- Fachbereich Wirtschaftswissenschaften (207)
- Solar-Institut Jülich (164)
- Fachbereich Bauingenieurwesen (153)
- ECSM European Center for Sustainable Mobility (79)
- MASKOR Institut für Mobile Autonome Systeme und Kognitive Robotik (67)
- Nowum-Energy (28)
- Fachbereich Gestaltung (25)
- Institut fuer Angewandte Polymerchemie (23)
- Sonstiges (21)
- Fachbereich Architektur (20)
- Freshman Institute (18)
- Kommission für Forschung und Entwicklung (18)
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.
Messenger apps like WhatsApp or Telegram are an integral part of daily communication. Besides the various positive effects, those services extend the operating range of criminals. Open trading groups with many thousand participants emerged on Telegram. Law enforcement agencies monitor suspicious users in such chat rooms. This research shows that text analysis, based on natural language processing, facilitates this through a meaningful domain overview and detailed investigations. We crawled a corpus from such self-proclaimed black markets and annotated five attribute types products, money, payment methods, user names, and locations. Based on each message a user sends, we extract and group these attributes to build profiles. Then, we build features to cluster the profiles. Pretrained word vectors yield better unsupervised clustering results than current
state-of-the-art transformer models. The result is a semantically meaningful high-level overview of the user landscape of black market chatrooms. Additionally, the extracted structured information serves as a foundation for further data exploration, for example, the most active users or preferred payment methods.
Purpose
To assess potential cognitive deficits under the influence of static magnetic fields at various field strengths some studies already exist. These studies were not focused on attention as the most vulnerable cognitive function. Additionally, mostly no magnetic resonance imaging (MRI) sequences were performed.
Materials and Methods
In all, 25 right-handed men were enrolled in this study. All subjects underwent one MRI examination of 63 minutes at 1.5 T and one at 7 T within an interval of 10 to 30 days. The order of the examinations was randomized. Subjects were referred to six standardized neuropsychological tests strictly focused on attention immediately before and after each MRI examination. Differences in neuropsychological variables between the timepoints before and after each MRI examination were assessed and P-values were calculated
Results
Only six subtests revealed significant differences between pre- and post-MRI. In these tests the subjects achieved better results in post-MRI testing than in pre-MRI testing (P = 0.013–0.032). The other tests revealed no significant results.
Conclusion
The improvement in post-MRI testing is only explicable as a result of learning effects. MRI examinations, even in ultrahigh-field scanners, do not seem to have any persisting influence on the attention networks of human cognition immediately after exposure.
A large strain collection comprising antagonistic bacteria was screened for novel detergent proteases. Several strains displayed protease activity on agar plates containing skim milk but were inactive in liquid media. Encapsulation of cells in alginate beads induced protease production. Stenotrophomonas maltophilia emerged as best performer under washing conditions. For identification of wash-active proteases, four extracellular serine proteases called StmPr1, StmPr2, StmPr3 and StmPr4 were cloned. StmPr2 and StmPr4 were sufficiently overexpressed in E. coli. Expression of StmPr1 and StmPr3 resulted in unprocessed, insoluble protein. Truncation of most of the C-terminal domain which has been identified by enzyme modeling succeeded in expression of soluble, active StmPr1 but failed in case of StmPr3.
From laundry application tests StmPr2 turned out to be a highly wash-active protease at 45 °C. Specific activity of StmPr2 determined with suc-l-Ala-l-Ala-l-Pro-l-Phe-p-nitroanilide as the substrate was 17 ± 2 U/mg. In addition we determined the kinetic parameters and cleavage preferences of protease StmPr2.
GaAs-based Gunn diodes with graded AlGaAs hot electron injector heterostructures have been developed under the special needs in automotive applications. The fabrication of the Gunn diode chips was based on total substrate removal and processing of integrated Au heat sinks. Especially, the thermal and RF behavior of the diodes have been analyzed by DC, impedance and S-parameter measurements. The electrical investigations have revealed the functionality of the hot electron injector. An optimized layer structure could fulfill the requirements in adaptive cruise control (ACC) systems at 77 GHz with typical output power between 50 and 90 mW.