Universität Stuttgart
Permanent URI for this communityhttps://elib.uni-stuttgart.de/handle/11682/1
Browse
Search Results
Item Open Access Using morpho-syntactic and semantic information to improve statistical machine translation(2018) Di Marco, Marion; Schulte im Walde, Sabine (PD Dr.)Statistische Maschinelle Übersetzungssystem werden von Wort-alignierten parallelen Corpora abgeleitet und benutzen üblicherweise keine expliziten linguistischen Informationen. Dies kann zu Generalisierungsproblemen führen, besonders wenn morphologisch komplexe Sprachen übersetzt werden. Diese Arbeit untersucht die Integration von linguistischen Informationen in ein Übersetzungssystem, das in eine morphologisch komplexe Sprache übersetzt: basierend auf einem Übersetzungssystem, das die Morphologie der Zielsprache modelliert, werden syntaktische und semantische Informationen in das System integriert, mit dem Ziel, die Modellierung von Subkategorisierung und Präpositionen zu verbessern.Item Open Access Computational approaches for German particle verbs: compositionality, sense discrimination and non-literal language(2018) Köper, Maximilian; Schulte im Walde, Sabine (PD Dr.)Anfangen (to start) is a German particle verb. Consisting of two parts, a base verb ("fangen") and particle ("an"), with potentially many or no intervening words in a sentence, particle verbs are highly frequent constructions with special properties. It has been shown that this type of verb represents a serious problem for language technology, due to particle verbs' ambiguity, ability to occur separate and seemingly unpredictable behaviour in terms of meaning. This dissertation addresses the meaning of German particle verbs via large-scale computational approaches. The three central parts of the thesis are concerned with computational models for the following components: i) compositionality, ii) senses and iii) non-literal language. In the first part of this thesis, we shed light on the phenomena by providing information on the properties of particle verbs, as well as the related and prior literature. In addition, we present the first corpus-driven statistical analysis. We use two different approaches for addressing the modelling of compositionality. For both approaches, we rely on large amounts of textual data with an algebraic model for representation to approximate meaning. We put forward the existing methodology and show that the prediction of compositionality can be improved by considering visual information. We model the particle verb senses based only on huge amounts of texts, without access to other resources. Furthermore, we compare and introduce the methods to find and represent different verb senses. Our findings indicate the usefulness of such sense-specific models. We successfully present the first model for detecting the non-literal language of particle verbs in a running text. Our approach reaches high performance by combining the established techniques from metaphor detection with particle verb-specific information. In the last part of the thesis, we approach the regularities and the meaning shift patterns. Here, we introduce a novel data collection approach for accessing the meaning components, as well as a computational model of particle verb analogy. The experiments reveal typical patterns in domain changes. Our data collection indicates that coherent verbs with the same meaning shift represent rather scarce phenomena. In summary, we provide novel computational models to previously unaddressed problems, and we report incremental improvements in the existing approaches. Across the models, we observe that semantically similar or synonymous base verbs behave similarly when combined with a particle. In addition, our models demonstrate the difficulty of particle verbs. Finally, our experiments suggest the usefulness of external normative emotion and affect ratings.Item Open Access Energieeffizienz von Prozessoren in High Performance Computinganwendungen der Ingenieurwissenschaften(Stuttgart : Höchstleistungsrechenzentrum, Universität Stuttgart, 2018) Khabi, Dmitry; Resch, Michael M. (Prof. Dr.-Ing. Dr. h.c. Dr. h.c. Prof. E.h.)Im Mittelpunkt dieser Arbeit steht die Frage nach Energieeffizienz im Hochleistungsrechnen (HPC) mit Schwerpunkt auf Zusammenhänge zwischen der elektrischen Leistung der Prozessoren und deren Rechenleistung. In Kapitel 1, Einleitung der folgenden Abhandlungen, werden die Motivation und der Stand der Technik auf dem Gebiet der Strommessung und der Energieeffizienz im HPC und dessen Komponenten erläutert. In den Folgenden Kapiteln 2 und 3 wird eine am Höchstleistungsrechenzentrum Stuttgart (HLRS) entwickelte Messtechnik detailliert diskutiert, die für die Strommessungen im Testcluster angewendet wird. Das Messverfahren der unterschiedlichen Hardwarekomponenten und die Abhängigkeit zwischen deren Stromversorgung, Messgenauigkeit und Messfrequenz werden dargelegt. Im Kapitel 4 der Arbeit beschreibe ich, welchen Zusammenhang es zwischen dem Stromverbrauch eines Prozessors, dessen Konfiguration und darauf ausgeführten Algorithmen gibt. Der Fokus liegt dabei auf den Zusammenhängen zwischen CPU-Frequenz, Grad der Parallelisierung, Rechenleistung und elektrischer Leistung. Für den Effizienzvergleich zwischen den Prozessoren und Algorithmen benutze ich ein Verfahren, das auf eine Approximation in der analytischen Form der Rechen- und der elektrischen Leistung der Prozessoren basiert. In diesem Kapitel wird außerdem gezeigt, dass die Koeffizienten der Approximation, die mehrere Hinweise auf Software und Hardware-Eigenschaften geben, als Basis für die Ausarbeitung eines erweiterten Modells dienen können. Wie im weiteren Verlauf gezeigt wird, berücksichtigen die existierenden Modelle der Rechen- und der elektrischen Leistung nur zum Teil die unterschiedlichen Frequenz-Domains der Hardwarekomponenten. Im Kapitel 5 wird eine Erweiterung des existierenden Modells der Rechenleistung erläutert, mit dessen Hilfe die entsprechenden neuen Eigenschaften der CPU-Architektur teilweise erklärt werden könnten. Die daraus gewonnenen Erkenntnisse sollen helfen, ein Modell zu entwickeln, das sowohl die Rechen- als auch die elektrische Leistung beschreibt. In Kapitel 6 beschreibe ich die Problemstellung der Energieeffizienz eines Hochleistungsrechners. Unter anderem werden die in dieser Arbeit entwickelten Methoden auf eine HPC-Platform evaluiert.Item Open Access Modell zum maschinellen Lernen von Wirkzusammenhängen bei der Holzverarbeitung auf Basis von online-erfassten Werkzeugmaschinendaten(Stuttgart : Fraunhofer Verlag, 2018) Lenz, Jürgen Herbert; Westkämper, Engelbert (Univ.-Prof. a. D. Dr.-Ing. Prof. E.h. Dr.-Ing. E.h. Dr. h.c. mult.)Aufgrund des immer härter werdenden globalen Wettbewerbs müssen produzierende Unternehmen, die auch in der Zukunft profitabel produzieren wollen, ihre Leistungsreserven nutzten. Die Möbelfertigung, die größte holzverarbeitende Industrie, besteht im Hauptprozess aus dem Fräsen von Holzwerkstoffen. Hierbei gibt es Leistungsreserven in der Einsatzplanung der Fräswerkzeuge. Gute Einsatzplanung ist die Voraussetzung für eine hohe Verfügbarkeit des Produktionssystems. Die Einsatzplanung wird durch Entwicklungen wie individuelle Möbelstücke, kleinere Losgrößen und neue Schneidstoffe erschwert. Die Herausforderung der Planungsunsicherheit beim Werkzeugeinsatz in der Holzbearbeitung wächst zusätzlich durch die größere Anzahl an industriell hergestellten Holzwerkstoffen mit jeweils unterschiedlicher Abrasivität. Dadurch wird die Bestimmung der Reststandzeit eines Werkzeuges erschwert. Zielsetzung dieser Arbeit ist die Planungssicherheit des Werkzeugeinsatzes durch eine exakte Planung des Werkzeugwechselfensters sowie durch Prognose der Reststandzeit zu erhöhen. Mithilfe dieser Prognose kann das gesamte Standvermögen des Werkzeuges verwendet werden. Das führt dazu, dass die Verfügbarkeit des Produktionssystems erhöht wird, da durch das Überschreiten der Werkzeugeinsatzgrenze bedingte Stillstände vermieden werden. Hierfür wurde ein Modell erstellt, das online erfasste Daten aus der Werkzeugmaschinensteuerung mit kontextbezogenen Informationen aus Datenbanken wie dem ERP-System und der Werkzeugverwaltung kombiniert. Aus diesen Informationen wird eine werkzeugspezifische Einsatzhistorie gebildet und mit gemessenen physikalischen Werten über den Werkzeugverschleiß und Kantenqualität des Werkstückes in Verbindung gebracht. Diese Verbindung von Bearbeitungshistorie und echten physikalischen Messgrößen bilden die Datenbasis für das maschinelle Lernen von Wirkzusammenhängen. Durch das Erlernen dieser Zusammenhänge kann die Reststandzeit eines Werkzeuges prognostiziert werden und somit die Planungsgenauigkeit des Werkzeugeinsatzes durch exakte Festlegung von Werkzeugwechselfenstern gesteigert werden. Zur Erprobung wurde das entwickelte Modell implementiert und seine Funktionsfähigkeit anhand einer Werkstoff-/Schneidstoffpaarung validiert. Diese Erprobung zeigte dass die Wirkzusammenhänge erlernt werden können.Item Open Access A massively parallel combination technique for the solution of high-dimensional PDEs(2018) Heene, Mario; Pflüger, Dirk (Jun.-Prof. Dr.)The solution of high-dimensional problems, especially high-dimensional partial differential equations (PDEs) that require the joint discretization of more than the usual three spatial dimensions and time, is one of the grand challenges in high performance computing (HPC). Due to the exponential growth of the number of unknowns - the so-called curse of dimensionality, it is in many cases not feasible to resolve the simulation domain as fine as required by the physical problem. Although the upcoming generation of exascale HPC systems theoretically provides the computational power to handle simulations that are out of reach today, it is expected that this is only achievable with new numerical algorithms that are able to efficiently exploit the massive parallelism of these systems. The sparse grid combination technique is a numerical scheme where the problem (e.g., a high-dimensional PDE) is solved on different coarse and anisotropic computational grids (so-called component grids), which are then combined to approximate the solution with a much higher target resolution than any of the individual component grids. This way, the total number of unknowns being computed is drastically reduced compared to the case when the problem is directly solved on a regular grid with the target resolution. Thus, the curse of dimensionality is mitigated. The combination technique is a promising approach to solve high-dimensional problems on future exascale systems. It offers two levels of parallelism: the component grids can be computed in parallel, independently and asynchronously of each other; and the computation of each component grid can be parallelized as well. This reduces the demand for global communication and synchronization, which is expected to be one of the limiting factors for classical discretization techniques to achieve scalability on exascale systems. Furthermore, the combination technique enables novel approaches to deal with the increasing fault rates expected from these systems. With the fault-tolerant combination technique it is possible to recover from failures without time-consuming checkpoint-restart mechanisms. In this work, new algorithms and data structures are presented that enable a massively parallel and fault-tolerant combination technique for time-dependent PDEs on large-scale HPC systems. The scalability of these algorithms is demonstrated on up to 180225 processor cores on the supercomputer Hazel Hen. Furthermore, the parallel combination technique is applied to gyrokinetic simulations in GENE, a software for the simulation of plasma microturbulence in fusion devices.Item Open Access Vision-based methods for evaluating visualizations(2018) Netzel, Rudolf; Weiskopf, Daniel (Prof. Dr.)Item Open Access Interactive web-based visualization(2018) Mwalongo, FinianThe visualization of large amounts of data, which cannot be easily copied for processing on a user’s local machine, is not yet a fully solved problem. Remote visualization represents one possible solution approach to the problem, and has long been an important research topic. Depending on the device used, modern hardware, such as high-performance GPUs, is sometimes not available. This is another reason for the use of remote visualization. Additionally, due to the growing global networking and collaboration among research groups, collaborative remote visualization solutions are becoming more important. The additional use of collaborative visualization solutions is eventually due to the growing global networking and collaboration among research groups. The attractiveness of web-based remote visualization is greatly increased by the wide availability of web browsers on almost all devices; these are available today on all systems - from desktop computers to smartphones. In order to ensure interactivity, network bandwidth and latency are the biggest challenges that web-based visualization algorithms have to solve. Despite the steady improvements in available bandwidth, these improvements are still significantly slower than, for example, processor performance, resulting in increasing the impact of this bottleneck. For example, visualization of large dynamic data in low-bandwidth environments can be challenging because it requires continuous data transfer. However, bandwidth improvement alone cannot improve the latency because it is also affected by factors such as the distance between server and client and network utilization. To overcome these challenges, a combination of techniques is needed to customize the individual processing steps of the visualization pipeline, from efficient data representation to hardware-accelerated rendering on the client side. This thesis first deals with related work in the field of remote visualization with a particular focus on interactive web-based visualization and then presents techniques for interactive visualization in the browser using modern web standards such as WebGL and HTML5. These techniques enable the visualization of dynamic molecular data sets with more than one million atoms at interactive frame rates using GPU-based ray casting. Due to the limitations which exist in a browser-based environment, the concrete implementation of the GPU-based ray casting had to be customized. Evaluation of the resulting performance shows that GPU-based techniques enable the interactive rendering of large data sets and achieve higher image quality compared to polygon-based techniques. In order to reduce data transfer times and network latency, and improve rendering speed, efficient approaches for data representation and transmission are used. Furthermore, this thesis introduces a GPU-based volume-ray marching technique based on WebGL 2.0, which uses progressive brick-wise data transfer, as well as multiple levels of detail in order to achieve interactive volume rendering of datasets stored on a server. The concepts and results presented in this thesis contribute to the further spread of interactive web-based visualization. The algorithmic and technological advances that have been achieved form a basis for further development of interactive browser-based visualization applications. At the same time, this approach has the potential for enabling future collaborative visualization in the cloud.Item Open Access Scheduling & routing time-triggered traffic in time-sensitive networks(2018) Nayak, Naresh Ganesh; Rothermel, Kurt (Prof. Dr. rer. nat. Dr. h. c.)The application of recent advances in computing, cognitive and networking technologies in manufacturing has triggered the so-called fourth industrial revolution, also referred to as Industry 4.0. Smart and flexible manufacturing systems are being conceived as a part of the Industry 4.0 initiative to meet the challenging requirements of the modern day manufacturers, e.g., production batch sizes of one. The information and communication technologies (ICT) infrastructure in such smart factories is expected to host heterogeneous applications ranging from the time-sensitive cyber-physical systems regulating physical processes in the manufacturing shopfloor to the soft real-time analytics applications predicting anomalies in the assembly line. Given the diverse demands of the applications, a single converged network providing different levels of communication guarantees to the applications based on their requirements is desired. Ethernet, on account of its ubiquity and its steadily growing performance along with shrinking costs, has emerged as a popular choice as a converged network. However, Ethernet networks, primarily designed for best-effort communication services, cannot provide strict guarantees like bounded end-to-end latency and jitter for real-time traffic without additional enhancements. Two major standardization bodies, viz., the IEEE Time-sensitive Networking (TSN) Task Group (TG) and the IETF Deterministic Networking (DetNets) Working Group are striving towards equipping Ethernet networks with mechanisms that would enable it to support different classes of real-time traffic. In this thesis, we focus on handling the time-triggered traffic (primarily periodic in nature) stemming from the hard real-time cyber-physical systems embedded in the manufacturing shopfloor over Ethernet networks. The basic approach for this is to schedule the transmissions of the time-triggered data streams appropriately through the network and ensure that the allocated schedules are adhered with. This approach leverages the possibility to precisely synchronize the clocks of the network participants, i.e., end systems and switches, using time synchronization protocols like the IEEE 1588 Precision Time Protocol (PTP). Based on the capabilities of the network participants, the responsibility of enforcing these schedules can be distributed. An important point to note is that the network utilization with respect to the time-triggered data streams depends on the computed schedules. Furthermore, the routing of the time-triggered data streams also influences the computed transmission schedules, and thus, affects the network utilization. The question however remains as to how to compute transmission schedules for time-triggered data streams along with their routes so that an optimal network utilization can be achieved. We explore, in this thesis, the scheduling and routing problems with respect to the time-triggered data streams in Ethernet networks. The recently published IEEE 802.1Qbv standard from the TSN-TG provides programmable gating mechanisms for the switches enabling them to schedule transmissions. Meanwhile, the extensions specified in the IEEE 802.1Qca standard or the primitives provided by OpenFlow, the popular southbound software-defined networking (SDN) protocol, can be used for gaining an explicit control over the routing of the data streams. Using these mechanisms, the responsibility of enforcing transmission schedules can be taken over by the end systems as well as the switches in the network. Alternatively, the scheduling can be enforced only by the end systems or only by the switches. Furthermore, routing alone can also be used to isolate time-triggered data streams, and thus, bound the latency and jitter experienced by the data streams in absence of synchronized clocks in the network. For each of the aforementioned cases, we formulate the scheduling and routing problem using Integer Linear Programming (ILP) for static as well as dynamic scenarios. The static scenario deals with the computation of schedules and routes for time-triggered data streams with a priori knowledge of their specifications. Here, we focus on computing schedules and routes that are optimal with respect to the network utilization. Given that the scheduling problems in the static setting have a high time-complexity, we also present efficient heuristics to approximate the optimal solution. With the dynamic scheduling problem, we address the modifications to the computed transmission schedules for adding further or removing already scheduled time-triggered data streams. Here, the focus lies on reducing the runtime of the scheduling and routing algorithms, and thus, have lower set-up times for adding new data streams into the network.Item Open Access Simulationsgestützte Absicherung von Fahrerassistenzsystemen(2018) Feilhauer, Marius; Resch, Michael M. (Prof. Dr.-Ing. Dr. h.c. Dr. h.c. Prof. E.h.)Item Open Access Interacting with large high-resolution display workplaces(2018) Lischke, Lars; Schmidt, Albrecht (Prof.)Large visual spaces provide a unique opportunity to communicate large and complex pieces of information; hence, they have been used for hundreds of years for varied content including maps, public notifications and artwork. Understanding and evaluating complex information will become a fundamental part of any office work. Large high-resolution displays (LHRDs) have the potential to further enhance the traditional advantages of large visual spaces and combine them with modern computing technology, thus becoming an essential tool for understanding and communicating data in future office environments. For successful deployment of LHRDs in office environments, well-suited interaction concepts are required. In this thesis, we build an understanding of how concepts for interaction with LHRDs in office environments could be designed. From the human-computer interaction (HCI) perspective three aspects are fundamental: (1) The way humans perceive and react to large visual spaces is essential for interaction with content displayed on LHRDs. (2) LHRDs require adequate input techniques. (3) The actual content requires well-designed graphical user interfaces (GUIs) and suitable input techniques. Perceptions influence how users can perform input on LHRD setups, which sets boundaries for the design of GUIs for LHRDs. Furthermore, the input technique has to be reflected in the design of the GUI. To understand how humans perceive and react to large visual information on LHRDs, we have focused on the influence of visual resolution and physical space. We show that increased visual resolution has an effect on the perceived media quality and the perceived effort and that humans can overview large visual spaces without being overwhelmed. When the display is wider than 2 m users perceive higher physical effort. When multiple users share an LHRD, they change their movement behavior depending whether a task is collaborative or competitive. For building LHRDs consideration must be given to the increased complexity of higher resolutions and physically large displays. Lower screen resolutions provide enough display quality to work efficiently, while larger physical spaces enable users to overview more content without being overwhelmed. To enhance user input on LHRDs in order to interact with large information pieces, we built working prototypes and analyzed their performance in controlled lab studies. We showed that eye-tracking based manual and gaze input cascaded (MAGIC) pointing can enhance target pointing to distant targets. MAGIC pointing is particularly beneficial when the interaction involves visual searches between pointing to targets. We contributed two gesture sets for mid-air interaction with window managers on LHRDs and found that gesture elicitation for an LHRD was not affected by legacy bias. We compared shared user input on an LHRD with personal tablets, which also functioned as a private working space, to collaborative data exploration using one input device together for interacting with an LHRD. The results showed that input with personal tablets lowered the perceived workload. Finally, we showed that variable movement resistance feedback enhanced one-dimensional data input when no visual input feedback was provided. We concluded that context-aware input techniques enhance the interaction with content displayed on an LHRD so it is essential to provide focus for the visual content and guidance for the user while performing input. To understand user expectations of working with LHRDs we prototyped with potential users how an LHRD work environment could be designed focusing on the physical screen alignment and the placement of content on the display. Based on previous work, we implemented novel alignment techniques for window management on LHRDs and compared them in a user study. The results show that users prefer techniques, that enhance the interaction without breaking well-known desktop GUI concepts. Finally, we provided the example of how an application for browsing scientific publications can benefit from extended display space. Overall, we show that GUIs for LHRDs should support the user more strongly than GUIs for smaller displays to arrange content meaningful or manage and understand large data sets, without breaking well-known GUI-metaphors. In conclusion, this thesis adopts a holistic approach to interaction with LHRDs in office environments. Based on enhanced knowledge about user perception of large visual spaces, we discuss novel input techniques for advanced user input on LHRDs. Furthermore, we present guidelines for designing future GUIs for LHRDs. Our work creates the design space of LHRD workplaces and identifies challenges and opportunities for the development of future office environments.