Showing papers by "Karlsruhe Institute of Technology published in 1998"

Knowledge engineering: principles and methods

Abstract: This paper gives an overview of the development of the field of Knowledge Engineering over the last 15 years. We discuss the paradigm shift from a transfer view to a modeling view and describe two approaches which considerably shaped research in Knowledge Engineering: Role-limiting Methods and Generic Tasks. To illustrate various concepts and methods which evolved in recent years we describe three modeling frameworks: CommonKADS, MIKE and PROTEGE-II. This description is supplemented by discussing some important methodological developments in more detail: specification languages for knowledge-based systems, problem-solving methods and ontologies. We conclude by outlining the relationship of Knowledge Engineering to Software Engineering, Information Integration and Knowledge Management.

Cold Bosonic Atoms in Optical Lattices

Abstract: The dynamics of an ultracold dilute gas of bosonic atoms in an optical lattice can be described by a Bose-Hubbard model where the system parameters are controlled by laser light We study the continuous (zero temperature) quantum phase transition from the superfluid to the Mott insulator phase induced by varying the depth of the optical potential, where the Mott insulator phase corresponds to a commensurate filling of the lattice (``optical crystal'') Examples for formation of Mott structures in optical lattices with a superimposed harmonic trap and in optical superlattices are presented

Early Stopping-But When?

Abstract: Validation can be used to detect when overfitting starts during supervised training of a neural network; training is then stopped before convergence to avoid the overfitting (“early stopping”). The exact criterion used for validation-based early stopping, however, is usually chosen in an ad-hoc fashion or training is stopped interactively. This trick describes how to select a stopping criterion in a systematic fashion; it is a trick for either speeding learning procedures or improving generalization, whichever is more important in the particular situation. An empirical investigation on multi-layer perceptrons shows that there exists a tradeoff between training time and generalization: From the given mix of 1296 training runs using different 12 problems and 24 different network architectures I conclude slower stopping criteria allow for small improvements in generalization (here: about 4% on average), but cost much more training time (here: about factor 4 longer on average).

Ordinary differential equations

Abstract: Many theorems of this chapter are very similar to the corresponding theorems of the preceding chapter. If we nevertheless devote a new chapter to an old subject, this will be justified by a new method. This method deals with differential equations and inequalities, whereas earlier the corresponding integral equations stood in the foreground. This is not done merely to give the old theorems a second proof. Rather, both of these first two chapters furnish, besides their concrete assertions, a methodology for later problems connected with partial differential equa­tions. Chapter III, devoted to hyperbolic differential equations, contains in essence a translation of Chapter I into several dimensions, while the theory of the parabolic equations in Chapter IV is closely connected to the present Chapter II.

Effects of increased solar ultraviolet radiation on terrestrial ecosystems

Abstract: Elevated solar UV-B radiation associated with stratospheric ozone reduction may exert effects on terrestrial ecosystems through actions on plants, microbes, and perhaps on some animals. At the ecosystem level, the effects are less well understood than at the molecular and organismal levels. Many of the most important, yet less predictable, consequences will be indirect effects of elevated UV-B acting through changes in the chemical composition and form of plants and through changes in the abiotic environment. These indirect effects include changes in the susceptibility of plants to attack by insects and pathogens in both agricultural and natural ecosystems; the direction of these changes can result in either a decrease or an increase in susceptibility. Other indirect effects of elevated UV-B include changes in competitive balance of plants and nutrient cycling. The direct UV-B action on plants that results in changes in form or function of plants appears to occur more often through altered gene activity rather than damage. The yield of some crop varieties can be decreased by elevated UV-B, but other varieties are not affected. Plant breeding and genetic engineering efforts should be able to cope with the potential threats to crop productivity due to elevated UV-B. For forest trees, this may be more difficult if effects of elevated UV-B accumulate over several years. All effects of elevated UV-B radiation must be considered in the context of other climate changes such as increased temperature and levels of carbon dioxide, which may alter the UV-B responses, especially for plants. The actions of elevated carbon dioxide and UV-B appear to be largely independent, but interactions occur between changes in UV-B and other factors. Other ecosystem-level consequences of elevated UV-B radiation are emerging and their magnitude and direction will not be easily predicted.

The signature of chemical valence in the electrical conduction through a single-atom contact

Abstract: Fabrication of structures at the atomic scale is now possible using state-of-the-art techniques for manipulating individual atoms1, and it may become possible to design electrical circuits atom by atom. A prerequisite for successful design is a knowledge of the relationship between the macroscopic electrical characteristics of such circuits and the quantum properties of the individual atoms used as building blocks. As a first step, we show here that the chemical valence determines the conduction properties of the simplest imaginable circuit—a one-atom contact between two metallic banks. The extended quantum states that carry the current from one bank to the other necessarily proceed through the valence orbitals of the constriction atom. It thus seems reasonable to conjecture that the number of current-carrying modes (or ‘channels’) of a one-atom contact is determined by the number of available valence orbitals, and so should strongly differ for metallic elements in different series of the periodic table. We have tested this conjecture using scanning tunnelling microscopy and mechanically controllable break-junction techniques2,3 to obtain atomic-size constrictions for four different metallic elements (Pb, Al, Nb and Au), covering a broad range of valences and orbital structures. Our results demonstrate unambiguously a direct link between valence orbitals and the number of conduction channels in one-atom contacts.

New Evidence against Hydroxyl Radicals as Reactive Intermediates in the Thermal and Photochemically Enhanced Fenton Reactions

Abstract: During the oxidative degradation of 2,4-dimethylaniline (2,4-xylidine) by means of the H2O2/UV method, a series of hydroxylated aromatic amines are formed, this result confirming the role of the hydroxyl radical as an initiator of the oxidative chain reaction. Thermal or photochemically enhanced Fenton reactions in the presence of 2,4-dimethylaniline (2,4-xylidine) yield primarily 2,4-dimethylphenol as an intermediate product, the genesis of which may only be explained by an electron transfer mechanism. Experimental evidence for such a mechanism is presented, and values for the quantum yields of the photochemically enhanced reduction of iron(III) to iron(II) in aqueous solutions of 2,4-xylidine are given.

Should computer scientists experiment more

Abstract: Computer scientists and practitioners defend their lack of experimentation with a wide range of arguments. Some arguments suggest that experimentation is inappropriate, too difficult, useless, and even harmful. This article discusses several such arguments to illustrate the importance of experimentation for computer science. It considers how the software industry is beginning to value experiments, because results may give a company a three- to five-year lead over the competition.

Performance of parallel TURBOMOLE for density functional calculations

Abstract: The parallelization of density functional treatments of molecular electronic energy and first-order gradients is described, and the performance is documented. The quadrature required for exchange correlation terms and the treatment of exact Coulomb interaction scales virtually linearly up to 100 nodes. The RI-J technique to approximate Coulomb interactions (by means of an auxiliary basis set approximation for the electron density) even shows superlinear speedup on distributed memory architectures. The bottleneck is then linear algebra. Demonstrative application examples include molecules with up to 300 atoms and 3000 basis functions that can now be treated in a few hours per geometry optimization cycle in C1 symmetry. © 1998 John Wiley & Sons, Inc. J Comput Chem 19: 1746–1757, 1998

Bosonization for beginners — refermionization for experts

Abstract: This tutorial review gives an elementary and self-contained derivation of the standard identities OwhOxUFh eif hOxU , etc.) for abelian bosonization in 1 dimension in a system of finite size L, following and simplifying Haldane's constructive approach. As a non-trivial application, we rigorously resolve (following Furusaki) a recent controversy regarding the tunneling density of states, r dos OwU, at the site of an impurity in a Tomonaga-Luttinger liquid: we use finite-size refer- mionization to show exactly that for ga 1 its asymptotic low-energy behavior is r dos OwUw. This agrees with the results of Fabrizio & Gogolin and of Furusaki, but not with those of Oreg and Finkel'stein (probably because we capture effects not included in their mean-field treatment of the Coulomb gas that they obtained by an exact mapping; their treatment of anti-commutation rela- tions in this mapping is correct, however, contrary to recent suggestions in the literature). — The tutorial is addressed to readers with little or no prior knowledge of bosonization, who are inter- ested in seeing ''all the details‚ explicitly; it is written at the level of beginning graduate students, requiring only knowledge of second quantization, but not of field theory (which is not needed here). At the same time, we hope that experts too might find useful our explicit treatment of certain subtleties that can often be swept under the rug, but are crucial for some applications, such as the calculation of rdosOwU - these include the proper treatment of the so-called Klein factors that act as fermion-number ladder operators (and also ensure the anti-commutation of different species of fermion fields), the retention of terms of order 1=L, and a novel, rigorous formulation of finite-size refermionization of both F eiFOxU and the boson field FOxU itself.

Web-based education for all: a tool for development adaptive courseware

Abstract: Many Web-based educational applications are expected to be used by very different groups of users without the assistance of a human teacher. Accordingly there is a need for systems which can adapt to users with very different backgrounds, prior knowledge of the subject and learning goals. An electronic textbook is one of the most prominent varieties of Web-based educational systems. In this paper we describe an approach for developing adaptive textbooks and present InterBook—an authoring tool based on this approach which simplifies the development of adaptive electronic textbooks on the Web.

Experiment versus Time Dependent Density Functional Theory Prediction of Fullerene Electronic Absorption

Abstract: Time dependent density functional theory (TDDFT) was applied to a description of optically allowed electronic transitions for C60 (Ih), C70 (D5h), C76 (D2), C78 (C2v), C78 (C2v‘), C78 (D3), C78 (D3h), C78 (D3h‘), and C80 (D2). We present a detailed comparison of experiment (solution spectra in toluene and n-hexane) and theory (vibrationless/gas phase) for all fullerenesexcept C78 (D3h, D3h‘) which have not yet been isolated. The level of agreement is good enough to allow assignment of dominant spectral features.

A systematic development of 'solid-shell' element formulations for linear and non-linear analyses employing only displacement degrees of freedom

Abstract: In the present contribution we propose a so-called solid-shell concept which incorporates only displacement degrees of freedom. Thus, some major disadvantages of the usually used degenerated shell concept are overcome. These disadvantages are related to boundary conditions—the handling of soft and hard support, the need for special co-ordinate systems at boundaries, the connection with continuum elements—and, in geometrically non-linear analyses, to a complicated update of the rotation vector. First, the kinematics of the so-called solid-shell concept in analogy to the degenerated shell concept are introduced. Then several modifications of the solid-shell concept are proposed to obtain locking-free solid-shell elements, leading also to formulations which allow the use of general three-dimensional material laws and which are also able to represent the normal stresses and strains in thickness direction. Numerical analyses of geometrically linear and non-linear problems are finally performed using solely assumed natural shear strain elements with a linear approximation in in-plane direction. Although some considerations are needed to get comparable boundary conditions in the examples analysed, the solid-shell elements prove to work as good as the degenerated shell elements. The numerical examples show that neither thickness nor shear locking are present even for distorted element shapes. © 1998 John Wiley & Sons, Ltd.

Tectonic shortening and crustal thickness in the Central Andes: How good is the correlation?

Abstract: Exceptional thickening of continental crust beneath the Central Andes is believed to be mainly a result of tectonic shortening of the South American plate in Neogene time. This shortening has been estimated to have contributed as much as 70%–80% of the present crustal volume. A compilation of published shortening values and our own estimates based on balanced cross sections from the Central Andes between lat 3°S and 40°S suggest that 70%–80% is a maximum rather than an average value for this part of the Andes. Tectonic shortening and the crustal cross-section area are only loosely correlated. Variations in tectonic shortening are more abrupt than those of crustal areas, particularly near the northern and southern ends of the Altiplano-Puna high plateau, where thick crust is associated with relatively small amounts of shortening. Shortening there may account for no more than about 30% of the present crustal cross-section area. The processes that created the remaining crustal area are not clear, but are likely to involve poorly constrained pre-Neogene tectonic shortening, moderate magmatic additions to the crust, tectonic underplating of material derived from the forearc, and possible flow of ductile lower crust along strike.

Effect of Dynamic Interfacial Tension on the Emulsification Process Using Microporous, Ceramic Membranes☆

Abstract: Conventional devices used in industrial emulsification processes disperse the inner phase by droplet disruption of high energetic laminar or turbulent flow. Membrane emulsification is different because small droplets are directly formed at the surface of a microporous membrane. Energy consumption of the process is lower, and the stresses on the system at the membrane surface and inside the pores are smaller. This allows processing of shear-sensitive substances. The result of the emulsification process can be described by the mean droplet size and the flux of the disperse phase. Among other parameters, pore size of the membrane, pressure of the disperse phase, and adsorption kinetics of the emulsifier influence the results of emulsification. The faster the emulsifier molecules adsorb at newly formed interfaces, the smaller the droplets of the emulsion produced. Transmembrane pressure greatly influences the flux but causes little change in droplet size.

The Stress Concept in Plants: An Introduction

Abstract: The current concept of stress in plants has been well developed over the past 60 years. Any unfavorable condition or substance that affects or blocks a plant's metabolism, growth, or development is regarded as stress. Vegetation stress can be induced by various natural and anthropogenic stress factors. One has to differentiate between short-term and long-term stress effects as well as between low-stress events that can be partially compensated for by acclimation, adaptation, and repair mechanisms, on the one hand, and strong stress or chronic stress events causing considerable damage that may eventually lead to cell and plant death, on the other hand. Some essential stress syndrome responses of plants are summarized in a unifying stress concept. The major abiotic, biotic, and anthropogenic stressors are listed. Some stress tolerance mechanisms are mentioned. Stress conditions and stress-induced damage in plants have so far been detected using the classical ecophysiological field methods as well as point data measurements of particular chlorophyll fluorescence parameters and of reflectance spectra. The novel laser-induced high-resolution fluorescence imaging technique, which integrates chlorophyll and blue-green fluorescence, marks a new standard in the detection of stress in plants.

Observation of the Bc meson in pp¯ collisions at √s = 1.8 TeV

Abstract: We have observed bottom-charm mesons via the decay mode B{sub c}{sup {+-}}{yields}J/{psi}{ell}{sup {+-}}v in 1.8 TeV p{ovr p}collisions using the CDF detector at the Fermilab Tevatron. A fit of background and signal contributions to the J/{psi}{ell} mass distribution yielded 20.4{sub -5.5}{sup +6.2} events from B{sub c} mesons. A fit to the same distribution with background alone was rejected at the level of 4.8 standard deviations. We measured the B{sub c}{sup +} mass to be 6.40{+-}0.39(stat){+-}0.13(syst) GeV/c{sup 2} and the B{sub c}{sup +} lifetime to be 0.46{sub -0.16}{sup +0.18}(stat){+-}0.03(syst) ps. Our measured yield (production cross section times branching ratio) for B{sub c}{sup +}{yields}J/{psi}{ell}{sup +}v relative to that for B{sup +}{yields}J/{psi} K{sup +} is 0.132{sub -0.037}{sup +0.041}(stat){+-}0.031(syst){sub -0.020}{sup +0.032}(lifetime).

Effect of ammonia on the anaerobic degradation of protein by a mesophilic and thermophilic biowaste population

Abstract: The influence of ammonia on the anaerobic degradation of peptone by mesophilic and thermophilic populations of biowaste was investigated. For peptone concentrations from 5 g l−1 to 20 g l−1 the mesophilic population revealed a higher rate of deamination than the thermophilic population, e.g. 552 mg l−1 day−1 compared to 320 mg l−1 day−1 at 10 g l−1 peptone. The final degree of deamination of the thermophilic population was, however, higher: 102 compared to 87 mg NH3/g peptone in the mesophilic cultures. If 0.5–6.5 g l−1 ammonia was added to the mesophilic biowaste cultures, deamination of peptone, degradation of its chemical oxygen demand (COD) and formation of biogas were increasingly inhibited, but no hydrogen was formed. The thermophilic biowaste cultures were most active if around 1 g ammonia l−1 was present. Deamination, COD degradation and biogas production decreased at lower and higher ammonia concentrations and hydrogen was formed in addition to methane. Studies of the inhibition by ammonia of peptone deamination, COD degradation and methane formation revealed a K i (50%) for NH3 of 92, 95 and 88 mg l−1 at 37 °C and 251, 274 and 297 mg l−1 at 55 °C respectively. This indicated that the thermophilic flora tolerated significantly more NH3 than the mesophilic flora. In the mesophilic reactor effluent 4.6 × 108 peptone-degrading colony-forming units (cfu)/ml were culturable, whereas in the thermophilic reactor effluent growth of only 5.6 × 107 cfu/ml was observed.

QCD corrections to the masses of the neutral CP -even Higgs bosons in the minimal supersymmetric standard model

Abstract: We perform a diagrammatic calculation of the leading two-loop QCD corrections to the masses of the neutral $\mathrm{CP}$-even Higgs bosons in the minimal supersymmetric standard model (MSSM). The results are valid for arbitrary values of the parameters of the Higgs boson and scalar top sector of the MSSM. The two-loop corrections are found to reduce the mass of the lightest Higgs boson considerably compared to its one-loop value. The numerical results are analyzed in the grand unified theory favored regions of small and large $\mathrm{tan}\ensuremath{\beta}.$ Their impact on a precise prediction for the mass of the lightest Higgs boson is briefly discussed.

Creating Robust Solutions by Means of Evolutionary Algorithms

Abstract: For real world problems it is often not sufficient to find solutions of high quality, but the solutions should also be robust. By robust we mean that the quality of the solution does not falter completely when a slight change of the environment occurs, or that certain deviations from the solution should be tolerated without a total loss of quality.