Zugl.: Leipzig, Univ., Diss., 2009

Zugl.: Würzburg, Univ., Diss., 2009

Universiẗat, Diss., 2004--München.

Zugl.: Regensburg, Universiẗat, Diss., 1998.

Techn. Universiẗat, Diss., 2004--Darmstadt.

Techn. University, Diss., 2004--Berlin.

University, Diss., 2005--Düsseldorf.

Background: Mathematical modelling of biological processes often requires a large variety of different data sets for parameter estimation and validation. It is common practice that clinical data are not available in raw formats but are provided as graphical representations. Hence, in order to include these data into environments used for model simulations and statistical analyses, it is necessary to extract them from their presentations in the literature. For this purpose, we developed the freely available open source tool ycasd. After establishing a coordinate system by simple axes definitions, it supports convenient retrieval of data points from arbitrary figures. Results: After describing the general functionality and providing an overview of the programme interface, we demonstrate on an example how to use ycasd. A major advantage of ycasd is that it does not require a certain input file format to open and process figures. All options of ycasd are accessible through a single window which eases handling and speeds up data extraction. For subsequent processing of extracted data points, results can be formatted as a Matlab or an R matrix. We extensively compare the functionality and other features of ycasd with other publically available tools. Finally, we provide a short summary of our experiences with ycasd in the context of modelling. Conclusions: We conclude that our tool is suitable for convenient and accurate data retrievals from graphical representations such as papers. Comparison of tools reveals that ycasd is a good compromise between easy and quick capturing of scientific data from publications and complexity. Our tool is routinely applied in the context of biological modelling, where numerous time series data are required to develop models. The software can also be useful for other kinds of analyses for which published data are required but are not available in raw formats such as systematic reviews and meta-analyses.

Background: Mathematical modelling of biological processes often requires a large variety of different data sets for parameter estimation and validation. It is common practice that clinical data are not available in raw formats but are provided as graphical representations. Hence, in order to include these data into environments used for model simulations and statistical analyses, it is necessary to extract them from their presentations in the literature. For this purpose, we developed the freely available open source tool ycasd. After establishing a coordinate system by simple axes definitions, it supports convenient retrieval of data points from arbitrary figures. Results: After describing the general functionality and providing an overview of the programme interface, we demonstrate on an example how to use ycasd. A major advantage of ycasd is that it does not require a certain input file format to open and process figures. All options of ycasd are accessible through a single window which eases handling and speeds up data extraction. For subsequent processing of extracted data points, results can be formatted as a Matlab or an R matrix. We extensively compare the functionality and other features of ycasd with other publically available tools. Finally, we provide a short summary of our experiences with ycasd in the context of modelling. Conclusions: We conclude that our tool is suitable for convenient and accurate data retrievals from graphical representations such as papers. Comparison of tools reveals that ycasd is a good compromise between easy and quick capturing of scientific data from publications and complexity. Our tool is routinely applied in the context of biological modelling, where numerous time series data are required to develop models. The software can also be useful for other kinds of analyses for which published data are required but are not available in raw formats such as systematic reviews and meta-analyses.

Based on the increased usage of heterogeneous information systems semistructural data models became more important in database technology. This thesis explains the basics of query processing in semistructural environments and discusses the possibilities of alternative access structures. In conjunction with these results two kinds of indexes are specified for the semistructural DBMS WebBus: path indexes and property indexes. These indexes are providing an efficient way for accessing data stored in the DBMS. / Bedingt durch die zunehmende Nutzung heterogener Informationssysteme gewinnt die semistrukturierte Datenmodellierung stetig an Bedeutung. Diese Arbeit setzt sich mit den Grundlagen der Anfrageverarbeitung in semistrukturierten Umgebungen auseinander und diskutiert die Möglichkeiten alternativer Zugriffspfade. Darauf aufbauend werden für das semistrukturierte Datenbank-Management-System WebBus Pfad- und Eigenschaftsindexe spezifiziert, die einen effizienten Zugriff auf die Daten des DBMS erlauben.