05 Fakultät Informatik, Elektrotechnik und Informationstechnik

Permanent URI for this collectionhttps://elib.uni-stuttgart.de/handle/11682/6

Browse

Filters

2004 - 200922010 - 20112

Settings

Institute: search.filters.UBSInstitut.Institut für Maschinelle SprachverarbeitungEnd date: 2011Subject: search.filters.subject.004

Search Results

Now showing 1 - 4 of 4
  • Thumbnail Image
    ItemOpen Access
    Methods for coreference visualization and annotation
    (2010) Burkovski, Andre; Heidemann, Gunther; Kobdani, Hamidreza; Schütze, Hinrich
    In this report we present methods for visualization of coreferences. We use Self Organizing Maps, graph-based models, and text-based techniques to visualize coreference information from a pair-wise model and offer an interactive presentation of the feature space. We introduce our software system and several visualizations of the Self Organizing Map. The visualizations enable the user to explore the feature space, provide insights for feature engineering and allow a fast annotation of data with coreference information.
  • Thumbnail Image
    ItemOpen Access
    Evaluation of automated business process optimization
    (2011) Ergin, Kemal Tolga
    Today's highly competitive markets tend to favor enterprises, in which business processes are analyzed and optimized regularly, in order to be able to operate in accordance with their business goals. The variety of business process management (BPM) methods applied for this purpose, since the emergence of the concept of business reengineering in the 1990s, ranges from incremental adjustments to radical restructuring. In combination with contemporary workflow automation technology, modern redesign methods are powerful tools for enhancing business performance, enabling companies to maintain a winning margin. Optimization methods that deliver sustainable results using evolutionary approaches, however, are nowadays becoming increasingly popular - once again, two decades after continuous improvement paradigms had almost completely been abandoned in favor of revolutionary process redesign. This diploma thesis explores one such evolutionary BPM approach employed in the deep Business Optimization Platform (dBOP), a research prototype, which assists analysts with the selection and application of suitable process improvement techniques. The present work demonstrates an evaluation of dBOP with the help of simulated business scenarios based on real case studies, and documents the types of optimization patterns most readily applied through automated process redesign. For this purpose two business processes, one from a car rental enterprise and one from a health insurance company, are modeled and deployed on a process server, and executed using web services and sample data warehouses based on actual statistics. These processes are then analyzed with dBOP, in order to compare its optimization recommendations with those expected from a human analyst's perspective.
  • Thumbnail Image
    ItemOpen Access
    Analysing names of organic chemical compounds : from morpho-semantics to SMILES strings and classes
    (2005) Anstein, Stefanie; Kremer, Gerhard
    The linguistic analysis of chemical terminology is a key to biochemical text processing and semi-automatic database curation. The system described analyses systematic and semi-systematic names of chemical compounds, class terms, and also otherwise underspecified names by means of a morpho-semantic grammar developed according to IUPAC nomenclature. It yields an intermediate semantic representation which describes the information encoded in a name. Our tool provides SMILES strings for the mapping of names to their molecule structure and also classifies the analysed terms. It was implemented in Prolog as a prototype and a basis for further development to support research in the life sciences.
  • Thumbnail Image
    ItemOpen Access
    Untersuchung der Sprecherindividualität höherer Formanten
    (2004) Kremer, Gerhard
    Die Behauptung, höhere Formanten seien sprecherspezifisch, scheint angemessen zu sein. Dennoch konnte ich keine Studien dazu finden, die dies belegen. In der vorliegenden Arbeit untersuchte ich die Individualität der Formanten F4-F7 der 16 Sprecher aus dem deutschen "Kiel-Korpus der spontanen gesprochenen Sprache". Auf der Grundlage einer LPC-Analyse ließ ich daraus mit Hilfe von Programmwerkzeugen von Entropic die Formantfrequenzdaten aller Sprachsignaldateien erzeugen. Mit einem Perl-Skript extrahierte ich die Daten der Vokale [a], [i], [u] und der Nasale [m], [n] und [\ng]. Ich wertete die Daten mit den statistischen Verfahren T-Test und ANOVA aus, wofür ich das Statistik-Programm R benutzte. Die Verteilungen der Formantfrequenzdaten unterschieden sich hochsignifikant bei allen sieben Formanten in Abhängigkeit der Faktoren Sprecher, Sprachlaut und der Sprecher-Sprachlaut-Kombination.