Refine
Year of publication
- 2019 (2589) (remove)
Document Type
- Article (1687)
- Doctoral Thesis (257)
- Postprint (211)
- Other (156)
- Monograph/Edited Volume (82)
- Review (68)
- Working Paper (36)
- Part of a Book (28)
- Part of Periodical (21)
- Master's Thesis (11)
- Conference Proceeding (10)
- Habilitation Thesis (7)
- Course Material (5)
- Bachelor Thesis (3)
- Report (3)
- Journal/Publication series (2)
- Contribution to a Periodical (1)
- Preprint (1)
Language
Keywords
- morphology (34)
- linguistics (31)
- syntax (31)
- Informationsstruktur (30)
- Morphologie (30)
- information structure (30)
- Festschrift (29)
- Linguistik (29)
- Syntax (29)
- festschrift (29)
Institute
- Institut für Biochemie und Biologie (352)
- Institut für Physik und Astronomie (312)
- Institut für Geowissenschaften (276)
- Institut für Chemie (183)
- Department Psychologie (110)
- Institut für Ernährungswissenschaft (89)
- Department Linguistik (86)
- Institut für Umweltwissenschaften und Geographie (80)
- Wirtschaftswissenschaften (73)
- Hasso-Plattner-Institut für Digital Engineering GmbH (69)
For a long time, there were things on this planet that only humans could do, but this time might be coming to an end. By using the universal tool that makes us unique – our intelligence – we have worked to eliminate our uniqueness, at least when it comes to solving cognitive tasks. Artificial intelligence is now able to play chess, understand language, and drive a car – and often better than we.
How did we get here? The philosopher Aristotle formulated the first “laws of thought” in his syllogisms, and the mathematicians Blaise Pascal and Wilhelm Leibniz built some of the earliest calculating machines. The mathematician George Boole was the first to introduce a formal language to represent logic. The natural scientist Alan Turing created his deciphering machine “Colossus,” the first programmable computer. Philosophers, mathematicians, psychologists, and linguists – for centuries, scientists have been developing formulas, machines, and theories that were supposed to enable us to reproduce and possibly even enhance our most valuable ability.
But what exactly is “artificial intelligence”? Even the name calls for comparison. Is artificial intelligence like human intelligence? Alan Turing came up with a test in 1950 to provide a satisfying operational definition of intelligence: According to him, a machine is intelligent if its thinking abilities equal those of humans. It has to reach human levels for any cognitive task. The machine has to prove this by convincing a human interrogator that it is human. Not an easy task: After all, it has to process natural language, store knowledge, draw conclusions, and learn something new. In fact, over the past ten years, a number of AI systems have emerged that have passed the test one way or another in chat conversations with automatically generated texts or images. Nowadays, the discussion usually centers on other questions: Does AI still need its creators? Will it not only outperform humans but someday replace them – be it in the world of work or even beyond? Will AI solve our problems in the age of all-encompassing digital networking – or will it become a part of the problem?
Artificial intelligence, its nature, its limitations, its potential, and its relationship to humans were being discussed even before it existed. Literature and film have created scenarios with very different endings. But what is the view of the scientists who are actually researching with or about artificial intelligence? For the current issue of our research magazine, a cognitive scientist, an education researcher, and a computer scientist shared their views. We also searched the University for projects whose professional environment reveals the numerous opportunities that AI offers for various disciplines. We cover the geosciences and computer science as well as economics, health, and literature studies.
At the same time, we have not lost sight of the broad research spectrum at the University: a legal expert introduces us to the not-so-distant sphere of space law while astrophysicists work on ensuring that state-of-the-art telescopes observe those regions in space where something “is happening” at the right time. A chemist explains why the battery of the future will come from a printer, and molecular biologists explain how they will breed stress-resistant plants. You will read about all this in this issue as well as about current studies on restless legs syndrome in children and the situation of Muslims in Brandenburg. Last but not least, we will introduce you to the sheep currently grazing in Sanssouci Park – all on behalf of science. Quite clever!
Enjoy your read!
THE EDITORS
Lange gab es auf der Erde Dinge, die konnte nur der Mensch. Doch diese Zeit könnte zu Ende gehen. Mithilfe des universalen Werkzeugs, das uns einzigartig macht – unserer Intelligenz –, haben wir dafür gesorgt, dass wir es nicht länger sind. Zumindest wenn es darum geht, kognitive Aufgaben zu lösen. Künstliche Intelligenz kann inzwischen Schach spielen, Sprache verstehen, Auto fahren. Vieles sogar besser als wir. Wie kam es dazu?
Der Philosoph Aristoteles schuf mit seinen Syllogismen die ersten „Gesetze des Denkens“, die Mathematiker Blaise Pascal und Wilhelm Leibniz bauten einige der frühesten Rechenmaschinen, der Mathematiker George Boole führte als erster eine formale Sprache zur Darstellung der Logik ein, der Naturwissenschaftler Alan Turing schuf mit seiner Dechiffriermaschine „Colossus“ den ersten programmierbaren Computer. Philosophen, Mathematiker, Psychologen, Linguisten – seit Jahrhunderten entwickeln Wissenschaftlerin- nen und Wissenschaftler Formeln, Maschinen und Theorien, die es möglich machen sollen, unsere wertvollste Fähigkeit zu reproduzieren und womöglich sogar zu verbessern. Aber was ist das eigentlich: „Künstliche Intelligenz“?
Schon die Bezeichnung fordert zum Vergleich auf. Ist Künstliche Intelligenz wie menschliche Intelligenz? Alan Turing formulierte 1950 einen Test, der eine befriedigende operationale Definition von Intelligenz liefern sollte: Intelligent ist eine Maschine demnach, wenn sie ein dem Menschen gleichwertiges Denkvermögen besitzt. Sie muss also bei beliebigen kognitiven Aufgaben dasselbe Niveau erreichen. Beweisen muss sie dies, indem sie einen menschlichen Fragenden glauben lässt, sie sei ein Mensch. Keine leichte Sache: Immerhin muss sie dafür natürliche Sprache verarbeiten, Wissen speichern, aus diesem Schlüsse ziehen und Neues lernen können. Tatsächlich entstanden in den vergangenen zehn Jahren etliche KI-Systeme, die in Chat- Gesprächen, mit automatisch erzeugten Texten oder Bildern den Test auf die eine oder andere Weise bestanden. Im Fokus stehen nun meist andere Fragen: Braucht KI ihre Schöpfer überhaupt noch? Wird sie den Menschen nicht nur überflügeln, sondern eines Tages sogar ersetzen – sei es in der Welt der Arbeit oder sogar darüber hinaus? Löst KI im Zeitalter der allumfassenden digitalen Vernetzung unsere Probleme – oder wird sie Teil davon?
Über Künstliche Intelligenz, ihr Wesen, ihre Beschränkungen, ihr Potenzial und ihr Verhältnis zum Menschen wird nicht erst diskutiert seitdem es sie gibt. Vor allem Literatur und Kino haben Szenarien mit verschiedenstem Ausgang kreiert. Aber wie sehen das Wissenschaftler, die mit oder zu Künstlicher Intelligenz forschen? Für die aktuelle Ausgabe des Forschungsmagazins kamen ein Kognitionswissenschaftler, eine Bildungsforscherin und ein Informatiker darüber ins Gespräch. Daneben haben wir uns in der Hochschule nach Projekten umgesehen, deren fachliche Heimat die zahlreichen Möglichkeiten offenbart, die KI für viele Disziplinen erahnen lässt. So geht die Reise in die Geowissenschaften und die Informatik ebenso wie die Wirtschafts-, Gesundheits- und Literaturwissenschaften.
Daneben haben wir die Breite der Forschung an der Universität nicht aus den Augen verloren: Ein Jurist führt ein in die gar nicht so weltferne Sphäre des Weltraumrechts, während Astrophysiker daran arbeiten, dass modernste Teleskope zum richtigen Zeitpunkt genau in die Regionen des Weltraums schauen, wo gerade etwas „los ist“. Eine Chemikerin erklärt, warum die Batterie der Zukunft aus dem Drucker kommt, und Molekularbiologen berichten, wie sie stressresistente Pflanzen züchten wollen. Mit menschlichem Stress in der Arbeitswelt beschäftigt sich nicht nur ein Forschungs-, sondern auch ein Gründerprojekt. Darüber ist in diesem Heft genauso zu lesen wie über aktuelle Studien zum Restless Legs Syndrom bei Kindern oder aber der Situation von Muslimen in Brandenburg. Nicht zuletzt machen wir Sie mit jenen Schafen bekannt, die derzeit im Park Sanssouci weiden – im Auftrag der Wissenschaft. Gar nicht so dumm! Viel Vergnügen!
Die Redaktion
Larix populations at the tundra-taiga ecotone in northern Siberia are highly under-represented in population genetic studies, possibly due to the remoteness of these regions that can only be accessed at extraordinary expense. The genetic signatures of populations in these boundary regions are therefore largely unknown. We aim to generate organelle reference genomes for the detection of single nucleotide polymorphisms (SNPs) that can be used for paleogenetic studies. We present 19 complete chloroplast genomes and mitochondrial genomic sequences of larches from the southern lowlands of the Taymyr Peninsula (northernmost range of Larix gmelinii (Rupr.) Kuzen.), the lower Omoloy River, and the lower Kolyma River (both in the range of Larix cajanderi Mayr). The genomic data reveal 84 chloroplast SNPs and 213 putatively mitochondrial SNPs. Parsimony-based chloroplast haplotype networks show no spatial structure of individuals from different geographic origins, while the mitochondrial haplotype network shows at least a slight spatial structure with haplotypes from the Omoloy and Kolyma populations being more closely related to each other than to most of the haplotypes from the Taymyr populations. Whole genome alignments with publicly available complete chloroplast genomes of different Larix species show that among official plant barcodes only the rcbL gene contains sufficient polymorphisms, but has to be sequenced completely to distinguish the different provenances. We provide 8 novel mitochondrial SNPs that are putatively diagnostic for the separation of L. gmelinii and L. cajanderi, while 4 chloroplast SNPs have the potential to distinguish the L. gmelinii/ L. cajanderi group from other Larix species. Our organelle references can be used for a targeted primer and probe design allowing the generation of short amplicons. This is particularly important with regard to future investigations of, for example, the biogeographic history of Larix by screening ancient sedimentary DNA of Larix.
Das Werk analysiert umfassend das Verbrechen der Aggression im Sinne des Römischen Statuts. Ausgehend von der Rechtsgeschichte, werde die einschlägigen Artikel 8bis, 15bis und 15ter des Römischen Statuts, also die Definition des Verbrechens der Aggression, analysiert.
Ebenso behandelt das Buch weiterführende Entwicklungen des Verbrechens der Aggression über das Jahr 2017 hinaus – das Jahr, in dem es, wahrscheinlich, zu einer Entscheidung über die Aktivierung der Gerichtsbarkeit kommt
Anti-Consumption
(2019)
Transcending the conventional debate around efficiency in sustainable consumption, anti-consumption patterns leading to decreased levels of material consumption have been gaining importance. Change agents are crucial for the promotion of such patterns, so there may be lessons for governance interventions that can be learnt from the every-day experiences of those who actively implement and promote sustainability in the field of anti-consumption. Eighteen social innovation pioneers, who engage in and diffuse practices of voluntary simplicity and collaborative consumption as sustainable options of anti-consumption share their knowledge and personal insights in expert interviews for this research. Our qualitative content analysis reveals drivers, barriers, and governance strategies to strengthen anti-consumption patterns, which are negotiated between the market, the state, and civil society. Recommendations derived from the interviews concern entrepreneurship, municipal infrastructures in support of local grassroots projects, regulative policy measures, more positive communication to strengthen the visibility of initiatives and emphasize individual benefits, establishing a sense of community, anti-consumer activism, and education. We argue for complementary action between top-down strategies, bottom-up initiatives, corporate activities, and consumer behavior. The results are valuable to researchers, activists, marketers, and policymakers who seek to enhance their understanding of materially reduced consumption patterns based on the real-life experiences of active pioneers in the field.
Transcending the conventional debate around efficiency in sustainable consumption, anti-consumption patterns leading to decreased levels of material consumption have been gaining importance. Change agents are crucial for the promotion of such patterns, so there may be lessons for governance interventions that can be learnt from the every-day experiences of those who actively implement and promote sustainability in the field of anti-consumption. Eighteen social innovation pioneers, who engage in and diffuse practices of voluntary simplicity and collaborative consumption as sustainable options of anti-consumption share their knowledge and personal insights in expert interviews for this research. Our qualitative content analysis reveals drivers, barriers, and governance strategies to strengthen anti-consumption patterns, which are negotiated between the market, the state, and civil society. Recommendations derived from the interviews concern entrepreneurship, municipal infrastructures in support of local grassroots projects, regulative policy measures, more positive communication to strengthen the visibility of initiatives and emphasize individual benefits, establishing a sense of community, anti-consumer activism, and education. We argue for complementary action between top-down strategies, bottom-up initiatives, corporate activities, and consumer behavior. The results are valuable to researchers, activists, marketers, and policymakers who seek to enhance their understanding of materially reduced consumption patterns based on the real-life experiences of active pioneers in the field.
New wave frequency and amplitude models for the nightside and dayside chorus waves are built based on measurements from the Electric and Magnetic Field Instrument Suite and Integrated Science (EMFISIS) instrument onboard the Van Allen Probes. The corresponding 3D diffusion coefficients are systematically obtained. Compared with previous commonly-used (typical) parameterizations, the new parameterizations result in differences in diffusion rates that depend on the energy and pitch angle. Furthermore, one-year 3D diffusive simulations are performed using the Versatile Electron Radiation Belt (VERB) code. Both typical and new wave parameterizations simulation results are in a good agreement with observations at 0.9 MeV. However, the new parameterizations for nightside chorus better reproduce the observed electron fluxes. These parameterizations will be incorporated into future modeling efforts.
In this paper we report a rare and fortunate event of fast magnetosonic (MS, also called equatorial noise) waves modulated by compressional ultralow frequency (ULF) waves measured by Van Allen Probes. The characteristics of MS waves, ULF waves, proton distribution, and their potential correlations are analyzed. The results show that ULF waves can modulate the energetic ring proton distribution and in turn modulate the MS generation. Furthermore, the variation of MS intensities is attributed to not only ULF wave activities but also the variation of background parameters, for example, number density. The results confirm the opinion that MS waves are generated by proton ring distribution and propose a new modulation phenomenon.
This study aims to identify the best-performing site characterization proxy alternative and complementary to the conventional 30 m average shear-wave velocity V-S30, as well as the optimal combination of proxies in characterizing linear site response. Investigated proxies include T-0 (site fundamental period obtained from earthquake horizontal-to-vertical spectral ratios), V-Sz (measured average shear-wave velocities to depth z, z = 5, 10, 20 and 30 m), Z(0.8) and Z(1.0) (measured site depths to layers having shear-wave velocity 0.8 and 1.0 km/s, respectively), as well as Z(x-infer) (inferred site depths from a regional velocity model, x = 0.8 and 1.0, 1.5 and 2.5 km/s). To evaluate the performance of a site proxy or a combination, a total of 1840 surface-borehole recordings is selected from KiK-net database. Site amplifications are derived using surface-to-borehole response-, Fourier- and cross-spectral ratio techniques and then are compared across approaches. Next, the efficacies of 7 single-proxies and 11 proxy-pairs are quantified based on the site-to-site standard deviation of amplification residuals of observation about prediction using the proxy or the pair. Our results show that T-0 is the best-performing single-proxy among T-0, Z(0.8), Z(1.0) and V-Sz. Meanwhile, T-0 is also the best-performing proxy among T-0, Z(0.8), Z(1.0) and Z(x-infer) complementary to V-S30 in accounting for the residual amplification after V-S30-correction. Besides, T-0 alone can capture most of the site effects and should be utilized as the primary site indicator. Though (T-0, V-S30) is the best-performing proxy pair among (V-S30, T-0), (V-S30, Z(0.8)), (V-S30, Z(1.0)), (V-S30, Z(x-infer)) and (T-0, V-Sz), it is only slightly better than (T-0, V-S20). Considering both efficacy and engineering utility, the combination of T-0 (primary) and V-S20 (secondary) is recommended. Further study is needed to test the performances of various proxies on sites in deep sedimentary basins.
In the Next Generation Attenuation West2 (NGA-West2) project, a 3D subsurface structure model (Japan Seismic Hazard Information Station [J-SHIS]) was queried to establish depths to 1.0 and 2.5 km/s velocity isosurfaces for sites without depth measurement in Japan. In this article, we evaluate the depth parameters in the J-SHIS velocity model by comparing them with their corresponding site-specific depth measurements derived from selected KiK-net velocity profiles. The comparison indicates that the J-SHIS model underestimates site depths at shallow sites and overestimates depths at deep sites. Similar issues were also identified in the southern California basin model. Our results also show that these underestimations and over-estimations have a potentially significant impact on ground-motion prediction using NGA-West2 ground-motion models (GMMs). Site resonant period may be considered as an alternative to depth parameter in the site term of a GMM.
We show that "stochastic bursting" is observed in a ring of unidirectional delay-coupled noisy excitable systems, thanks to the combinational action of time-delayed coupling and noise. Under the approximation of timescale separation, i.e., when the time delays in each connection are much larger than the characteristic duration of the spikes, the observed rather coherent spike pattern can be described by an idealized coupled point processwith a leader-follower relationship. We derive analytically the statistics of the spikes in each unit, the pairwise correlations between any two units, and the spectrum of the total output from the network. Theory is in good agreement with the simulations with a network of theta-neurons. Published under license by AIP Publishing.
The Kp index is a measure of the midlatitude global geomagnetic activity and represents short-term magnetic variations driven by solar wind plasma and interplanetary magnetic field. The Kp index is one of the most widely used indicators for space weather alerts and serves as input to various models, such as for the thermosphere and the radiation belts. It is therefore crucial to predict the Kp index accurately. Previous work in this area has mostly employed artificial neural networks to nowcast Kp, based their inferences on the recent history of Kp and on solar wind measurements at L1. In this study, we systematically test how different machine learning techniques perform on the task of nowcasting and forecasting Kp for prediction horizons of up to 12 hr. Additionally, we investigate different methods of machine learning and information theory for selecting the optimal inputs to a predictive model. We illustrate how these methods can be applied to select the most important inputs to a predictive model of Kp and to significantly reduce input dimensionality. We compare our best performing models based on a reduced set of optimal inputs with the existing models of Kp, using different test intervals, and show how this selection can affect model performance.
RNA-based processes play key roles in the regulation of eukaryotic gene expression. This includes both the processing of pre-mRNAs into mature mRNAs ready for translation and RNA-based silencing processes, such as RNA-directed DNA methylation (RdDM). Polyadenylation of pre-mRNAs is one important step in their processing and is carried out by three functionally specialized canonical nuclear poly(A) polymerases in Arabidopsis thaliana. Null mutations in one of these, termed PAPS1, result in a male gametophytic defect. Using a fluorescence-labelling strategy, we have characterized this defect in more detail using RNA and small-RNA sequencing. In addition to global defects in the expression of pollen-differentiation genes, paps1 null-mutant pollen shows a strong overaccumulation of transposable element (TE) transcripts, yet a depletion of 21- and particularly 24-nucleotide-long short interfering RNAs (siRNAs) and microRNAs (miRNAs) targeting the corresponding TEs. Double-mutant analyses support a specific functional interaction between PAPS1 and components of the RdDM pathway, as evident from strong synergistic phenotypes in mutant combinations involving paps1, but not paps2 paps4, mutations. In particular, the double-mutant of paps1 and rna-dependent rna polymerase 6 (rdr6) shows a synergistic developmental phenotype disrupting the formation of the transmitting tract in the female gynoecium. Thus, our findings in A. thaliana uncover a potentially general link between canonical poly(A) polymerases as components of mRNA processing and RdDM, reflecting an analogous interaction in fission yeast.
Paleogeographic reconstructions of terranes can greatly benefit from the provenance analysis of sediments. A series of Cenozoic basins provide key sedimentary archives for investigating the growth of the Tibetan Plateau, yet the provenance of the sediments in these basins has never been constrained robustly. Here we report sedimentary petrological and detrital zircon geochronological data from the Paleocene-Eocene Nangqian-Xialaxiu and Gongjue basins. Sandstone detrital modes and zircon morphology suggest that the samples collected in these two basins were sourced from recycled orogen. Detrital zircon geochronology indicates that sediments in the Nangqian-Xialaxiu Basin are characterized by two distinct age populations at 220-280 Ma and 405-445 Ma. In contrast, three predominant age populations of 207-256 Ma, 423-445 Ma, and 1851-1868 Ma, and two subordinate age populations of similar to 50 Ma and similar to 2500 Ma, are recognized in the Gongjue Basin. Comparison with detrital zircon ages from the surrounding terranes suggests that sediments in the Nangqian-Xialaxiu Basin come from the neighboring thrust belts, whereas sediments from the Gongjue Basin are predominantly derived from the distant Songpan-Ganzi Terrane with minor contribution from the surrounding areas. A three-stage Cenozoic evolution of the eastern Tibetan Plateau is proposed. During the Paleocene, the Nangqian-Xialaxiu Basin appeared as a set of small intermontane sub-basins and received plentiful sediments from the neighboring mountain belts; during the Eocene, the Gongjue Basin kept a relatively low altitude and was a depression at the edge of a proto-Plateau; since the Oligocene, the Tibetan Plateau further uplifted and the marginal Gongjue Basin was involved in the Tibetan interior orogeny, indicating the eastward propagation of the Tibetan Plateau.
Electrosynthesis and characterization of molecularly imprinted polymers for peptides and proteins
(2019)
Exploration of metal clusters (MCs) adaptive to both aqueous and oil phases without disturbing their size is promising for a broad scope of applications. The state-of-the-art approach via ligand-binding may perturb MCs' size due to varied metal–ligand binding strength when shuttling between solvents of different polarity. Herein, we applied physical confinement of a series of small noble MCs (<1 nm) inside ionic organic cages (I-Cages), which by means of anion exchange enables reversible transfer of MCs between aqueous and hydrophobic solutions without varying their ultrasmall size. Moreover, the MCs@I-Cage hybrid serves as a recyclable, reaction-switchable catalyst featuring high activity in liquid-phase NH3BH3 (AB) hydrolysis reaction with a turnover frequency (TOF) of 115 min−1.