Doctoral Thesis
Refine
Year of publication
Document Type
- Doctoral Thesis (29) (remove)
Language
- English (29) (remove)
Keywords
- Software Engineering (3)
- Modellgetriebene Entwicklung (2)
- Semantic Web (2)
- AUTOSAR (1)
- Abduktion <Logik> (1)
- Ad-hoc-Netz (1)
- Adaptation (1)
- Algorithmische Geometrie (1)
- Anpassung (1)
- Architektur <Informatik> (1)
Institute
- Institut für Informatik (29) (remove)
Web 2.0 provides technologies for online collaboration of users as well as the creation, publication and sharing of user-generated contents in an interactive way. Twitter, CNET, CiteSeerX, etc. are examples of Web 2.0 platforms which facilitate users in these activities and are viewed as rich sources of information. In the platforms mentioned as examples, users can participate in discussions, comment others, provide feedback on various issues, publish articles and write blogs, thereby producing a high volume of unstructured data which at the same time leads to an information overload. To satisfy various types of human information needs arising from the purpose and nature of the platforms requires methods for appropriate aggregation and automatic analysis of this unstructured data. In this thesis, we propose methods which attempt to overcome the problem of information overload and help in satisfying user information needs in three scenarios.
To this end, first we look at two of the main challenges of sparsity and content quality in Twitter and how these challenges can influence standard retrieval models. We analyze and identify Twitter content features that reflect high quality information. Based on this analysis we introduce the concept of "interestingness" as a static quality measure. We empirically show that our proposed measure helps in retrieving and filtering high quality information in Twitter. Our second contribution relates to the content diversification problem in a collaborative social environment, where the motive of the end user is to gain a comprehensive overview of the pros and cons of a discussion track which results from social collaboration of the people. For this purpose, we develop the FREuD approach which aims at solving the content diversification problem by combining latent semantic analysis with sentiment estimation approaches. Our evaluation results show that the FREuD approach provides a representative overview of sub-topics and aspects of discussions, characteristic user sentiments under different aspects, and reasons expressed by different opponents. Our third contribution presents a novel probabilistic Author-Topic-Time model, which aims at mining topical trends and user interests from social media. Our approach solves this problem by means of Bayesian modeling of relations between authors, latent topics and temporal information. We present results of application of the model to the scientific publication datasets from CiteSeerX showing improved semantically cohesive topic detection and capturing shifts in authors" interest in relation to topic evolution.
This thesis addresses the problem of terrain classification in unstructured outdoor environments. Terrain classification includes the detection of obstacles and passable areas as well as the analysis of ground surfaces. A 3D laser range finder is used as primary sensor for perceiving the surroundings of the robot. First of all, a grid structure is introduced for data reduction. The chosen data representation allows for multi-sensor integration, e.g., cameras for color and texture information or further laser range finders for improved data density. Subsequently, features are computed for each terrain cell within the grid. Classification is performedrnwith a Markov random field for context-sensitivity and to compensate for sensor noise and varying data density within the grid. A Gibbs sampler is used for optimization and is parallelized on the CPU and GPU in order to achieve real-time performance. Dynamic obstacles are detected and tracked using different state-of-the-art approaches. The resulting information - where other traffic participants move and are going to move to - is used to perform inference in regions where the terrain surface is partially or completely invisible for the sensors. Algorithms are tested and validated on different autonomous robot platforms and the evaluation is carried out with human-annotated ground truth maps of millions of measurements. The terrain classification approach of this thesis proved reliable in all real-time scenarios and domains and yielded new insights. Furthermore, if combined with a path planning algorithm, it enables full autonomy for all kinds of wheeled outdoor robots in natural outdoor environments.
The publication of freely available and machine-readable information has increased significantly in the last years. Especially the Linked Data initiative has been receiving a lot of attention. Linked Data is based on the Resource Description Framework (RDF) and anybody can simply publish their data in RDF and link it to other datasets. The structure is similar to the World Wide Web where individual HTML documents are connected with links. Linked Data entities are identified by URIs which are dereferenceable to retrieve information describing the entity. Additionally, so called SPARQL endpoints can be used to access the data with an algebraic query language (SPARQL) similar to SQL. By integrating multiple SPARQL endpoints it is possible to create a federation of distributed RDF data sources which acts like one big data store.
In contrast to the federation of classical relational database systems there are some differences for federated RDF data. RDF stores are accessed either via SPARQL endpoints or by resolving URIs. There is no coordination between RDF data sources and machine-readable meta data about a source- data is commonly limited or not available at all. Moreover, there is no common directory which can be used to discover RDF data sources or ask for sources which offer specific data. The federation of distributed and linked RDF data sources has to deal with various challenges. In order to distribute queries automatically, suitable data sources have to be selected based on query details and information that is available about the data sources. Furthermore, the minimization of query execution time requires optimization techniques that take into account the execution cost for query operators and the network communication overhead for contacting individual data sources. In this thesis, solutions for these problems are discussed. Moreover, SPLENDID is presented, a new federation infrastructure for distributed RDF data sources which uses optimization techniques based on statistical information.
Traditional Driver Assistance Systems (DAS) like for example Lane Departure Warning Systems or the well-known Electronic Stability Program have in common that their system and software architecture is static. This means that neither the number and topology of Electronic Control Units (ECUs) nor the presence and functionality of software modules changes after the vehicles leave the factory.
However, some future DAS do face changes at runtime. This is true for example for truck and trailer DAS as their hardware components and software entities are spread over both parts of the combination. These new requirements cannot be faced by state-of-the-art approaches of automotive software systems. Instead, a different technique of designing such Distributed Driver Assistance Systems (DDAS) needs to be developed. The main contribution of this thesis is the development of a novel software and system architecture for dynamically changing DAS using the example of driving assistance for truck and trailer. This architecture has to be able to autonomously detect and handle changes within the topology. In order to do so, the system decides which degree of assistance and which types of HMI can be offered every time a trailer is connected or disconnected. Therefore an analysis of the available software and hardware components as well as a determination of possible assistance functionality and a re-configuration of the system take place. Such adaptation can be granted by the principles of Service-oriented Architecture (SOA). In this architectural style all functionality is encapsulated in self-contained units, so-called Services. These Services offer the functionality through well-defined interfaces whose behavior is described in contracts. Using these Services, large-scale applications can be built and adapted at runtime. This thesis describes the research conducted in achieving the goals described by introducing Service-oriented Architectures into the automotive domain. SOA deals with the high degree of distribution, the demand for re-usability and the heterogeneity of the needed components.
It also applies automatic re-configuration in the event of a system change. Instead of adapting one of the frameworks available to this scenario, the main principles of Service-orientation are picked up and tailored. This leads to the development of the Service-oriented Driver Assistance (SODA) framework, which implements the benefits of Service-orientation while ensuring compatibility and compliance to automotive requirements, best-practices and standards. Within this thesis several state-of-the-art Service-oriented frameworks are analyzed and compared. Furthermore, the SODA framework as well as all its different aspects regarding the automotive software domain are described in detail. These aspects include a well-defined reference model that introduces and relates terms and concepts and defines an architectural blueprint. Furthermore, some of the modules of this blueprint such as the re-configuration module and the Communication Model are presented in full detail. In order to prove the compliance of the framework regarding state-of-the-art automotive software systems, a development process respecting today's best practices in automotive design procedures as well as the integration of SODA into the AUTOSAR standard are discussed. Finally, the SODA framework is used to build a full-scale demonstrator in order to evaluate its performance and efficiency.
One of the main goals of the artificial intelligence community is to create machines able to reason with dynamically changing knowledge. To achieve this goal, a multitude of different problems have to be solved, of which many have been addressed in the various sub-disciplines of artificial intelligence, like automated reasoning and machine learning. The thesis at hand focuses on the automated reasoning aspects of these problems and address two of the problems which have to be overcome to reach the afore-mentioned goal, namely 1. the fact that reasoning in logical knowledge bases is intractable and 2. the fact that applying changes to formalized knowledge can easily introduce inconsistencies, which leads to unwanted results in most scenarios.
To ease the intractability of logical reasoning, I suggest to adapt a technique called knowledge compilation, known from propositional logic, to description logic knowledge bases. The basic idea of this technique is to compile the given knowledge base into a normal form which allows to answer queries efficiently. This compilation step is very expensive but has to be performed only once and as soon as the result of this step is used to answer many queries, the expensive compilation step gets worthwhile. In the thesis at hand, I develop a normal form, called linkless normal form, suitable for knowledge compilation for description logic knowledge bases. From a computational point of view, the linkless normal form has very nice properties which are introduced in this thesis.
For the second problem, I focus on changes occurring on the instance level of description logic knowledge bases. I introduce three change operators interesting for these knowledge bases, namely deletion and insertion of assertions as well as repair of inconsistent instance bases. These change operators are defined such that in all three cases, the resulting knowledge base is ensured to be consistent and changes performed to the knowledge base are minimal. This allows us to preserve as much of the original knowledge base as possible. Furthermore, I show how these changes can be applied by using a transformation of the knowledge base.
For both issues I suggest to adapt techniques successfully used in other logics to get promising methods for description logic knowledge bases.
Reactive local algorithms are distributed algorithms which suit the needs of battery-powered, large-scale wireless ad hoc and sensor networks particularly well. By avoiding both unnecessary wireless transmissions and proactive maintenance of neighborhood tables (i.e., beaconing), such algorithms minimize communication load and overhead, and scale well with increasing network size. This way, resources such as bandwidth and energy are saved, and the probability of message collisions is reduced, which leads to an increase in the packet reception ratio and a decrease of latencies.
Currently, the two main application areas of this algorithm type are geographic routing and topology control, in particular the construction of a node's adjacency in a connected, planar representation of the network graph. Geographic routing enables wireless multi-hop communication in the absence of any network infrastructure, based on geographic node positions. The construction of planar topologies is a requirement for efficient, local solutions for a variety of algorithmic problems.
This thesis contributes to reactive algorithm research in two ways, on an abstract level, as well as by the introduction of novel algorithms:
For the very first time, reactive algorithms are considered as a whole and as an individual research area. A comprehensive survey of the literature is given which lists and classifies known algorithms, techniques, and application domains. Moreover, the mathematical concept of O- and Omega-reactive local topology control is introduced. This concept unambiguously distinguishes reactive from conventional, beacon-based, topology control algorithms, serves as a taxonomy for existing and prospective algorithms of this kind, and facilitates in-depth investigations of the principal power of the reactive approach, beyond analysis of concrete algorithms.
Novel reactive local topology control and geographic routing algorithms are introduced under both the unit disk and quasi unit disk graph model. These algorithms compute a node's local view on connected, planar, constant stretch Euclidean and topological spanners of the underlying network graph and route messages reactively on these spanners while guaranteeing the messages' delivery. All previously known algorithms are either not reactive, or do not provide constant Euclidean and topological stretch properties. A particularly important partial result of this work is that the partial Delaunay triangulation (PDT) is a constant stretch Euclidean spanner for the unit disk graph.
To conclude, this thesis provides a basis for structured and substantial research in this field and shows the reactive approach to be a powerful tool for algorithm design in wireless ad hoc and sensor networking.
Confidentiality, integrity, and availability are often listed as the three major requirements for achieving data security and are collectively referred to as the C-I-A triad. Confidentiality of data restricts the data access to authorized parties only, integrity means that the data can only be modified by authorized parties, and availability states that the data must always be accessible when requested. Although these requirements are relevant for any computer system, they are especially important in open and distributed networks. Such networks are able to store large amounts of data without having a single entity in control of ensuring the data's security. The Semantic Web applies to these characteristics as well as it aims at creating a global and decentralized network of machine-readable data. Ensuring the confidentiality, integrity, and availability of this data is therefore also important and must be achieved by corresponding security mechanisms. However, the current reference architecture of the Semantic Web does not define any particular security mechanism yet which implements these requirements. Instead, it only contains a rather abstract representation of security.
This thesis fills this gap by introducing three different security mechanisms for each of the identified security requirements confidentiality, integrity, and availability of Semantic Web data. The mechanisms are not restricted to the very basics of implementing each of the requirements and provide additional features as well. Confidentiality is usually achieved with data encryption. This thesis not only provides an approach for encrypting Semantic Web data, it also allows to search in the resulting ciphertext data without decrypting it first. Integrity of data is typically implemented with digital signatures. Instead of defining a single signature algorithm, this thesis defines a formal framework for signing arbitrary Semantic Web graphs which can be configured with various algorithms to achieve different features. Availability is generally supported by redundant data storage. This thesis expands the classical definition of availability to compliant availability which means that data must only be available as long as the access request complies with a set of predefined policies. This requirement is implemented with a modular and extensible policy language for regulating information flow control. This thesis presents each of these three security mechanisms in detail, evaluates them against a set of requirements, and compares them with the state of the art and related work.
The provision of electronic participation services (e-participation) is a complex socio-technical undertaking that needs comprehensive design and implementation strategies. E-participation service providers, in the most cases administrations and governments, struggle with changing requirements that demand more transparency, better connectivity and increased collaboration among different actors. At the same time, less staff are available. As a result, recent research assesses only a minority of e-participation services as successful. The challenge is that the e-participation domain lacks comprehensive approaches to design and implement (e-)participation services. Enterprise Architecture (EA) frameworks have evolved in information systems research as an approach to guide the development of complex socio-technical systems. This approach can guide the design and implementation services, if the collection of organisations with the commonly held goal to provide participation services is understood as an E Participation Enterprise (EE). However, research & practice in the e participation domain has not yet exploited EA frameworks. Consequently, the problem scope that motivates this dissertation is the existing gap in research to deploy EA frameworks in e participation design and implementation. The research question that drives this research is: What methodical and technical guides do architecture frameworks provide that can be used to design and implement better and successful e participation?
This dissertation presents a literature study showing that existing approaches have not covered yet the challenges of comprehensive e participation design and implementation. Accordingly, the research moves on to investigate established EA frameworks such as the Zachman Framework, TOGAF, the DoDAF, the FEA, the ARIS, and the ArchiMate for their use. While the application of these frameworks in e participation design and implementation is possible, an integrated approach is lacking so far. The synthesis of literature review and practical insights in design and implementation of e participation services from four projects show the challenges of adapting architecture frameworks for this domain. However, the research shows also the potential of a combination of the different approaches. Consequently, the research moves on to develop the E-Participation Architecture Framework (EPART-Framework). Therefore, the dissertation applies design science research including literature review and action research. Two independent settings test an initial EPART-Framework version. The results yield into the EPART-Framework presented in this dissertation.
The EPART-Framework comprises of the EPART-Metamodel with six EPART-Viewpoints, which frame the stakeholder concerns: the Participation Scope, the Participant Viewpoint, the Participation Viewpoint, the Data & Information Viewpoint, the E-participation Viewpoint, and Implementation & Governance Viewpoint. The EPART-Method supports the stakeholders to design the EE and implement e participation and stores its output in an architecture description and a solution repository. It consists of five consecutive phases accompanied by requirements management: Initiation, Design, Implementation and Preparation, Participation, and Evaluation. The EPART-Framework fills the gap between the e participation domain and the enterprise architecture framework domain. The evaluation gives reasonable evidence that the framework is a valuable addition in academia and in practice to improve e-participation design and implementation. The same time, it shows opportunities for future research to extend and advance the framework.
The term “Software Chrestomaty” is defined as a collection of software systems meant to be useful in learning about or gaining insight into software languages, software technologies, software concepts, programming, and software engineering. 101companies software chrestomathy is a community project with the attributes of a Research 2.0 infrastructure for various stakeholders in software languages and technology communities. The core of 101companies combines a semantic wiki and confederated open source repositories. We designed and developed an integrated ontology-based knowledge base about software languages and technologies. The knowledge is created by the community of contributors and supported with a running example and structured documentation. The complete ecosystem is exposed by using Linked Data principles and equipped with the additional metadata about individual artifacts. Within the context of software chrestomathy we explored a new type of software architecture – linguistic architecture that is targeted on the language and technology relationships within a software product and based on the megamodels. Our approach to documentation of the software systems is highly structured and makes use of the concepts of the newly developed megamodeling language MegaL. We “connect” an emerging ontology with the megamodeling artifacts to raise the cognitive value of the linguistic architecture.
Initial goal of the current dissertation was the determination of image-based biomarkers sensitive for neurodegenerative processes in the human brain. One such process is the demyelination of neural cells characteristic for Multiple sclerosis (MS) - the most common neurological disease in young adults for which there is no cure yet. Conventional MRI techniques are very effective in localizing areas of brain tissue damage and are thus a reliable tool for the initial MS diagnosis. However, a mismatch between the clinical fndings and the visualized areas of damage is observed, which renders the use of the standard MRI diffcult for the objective disease monitoring and therapy evaluation. To address this problem, a novel algorithm for the fast mapping of myelin water content using standard multiecho gradient echo acquisitions of the human brain is developed in the current work. The method extents a previously published approach for the simultaneous measurement of brain T1, T∗ 2 and total water content. Employing the multiexponential T∗ 2 decay signal of myelinated tissue, myelin water content is measured based on the quantifcation of two water pools (myelin water and rest) with different relaxation times. Whole brain in vivo myelin water content maps are acquired in 10 healthy controls and one subject with MS. The in vivo results obtained are consistent with previous reports. The acquired quantitative data have a high potential in the context of MS. However, the parameters estimated in a multiparametric acquisition are correlated and constitute therefore an ill-posed, nontrivial data analysis problem. Motivated by this specific problem, a new data clustering approach is developed called Nuclear Potential Clustering, NPC. It is suitable for the explorative analysis of arbitrary dimensional and possibly correlated data without a priori assumptions about its structure. The developed algorithm is based on a concept adapted from nuclear physics. To partition the data, the dynamic behavior of electrically even charged nucleons interacting in a d-dimensional feature space is modeled. An adaptive nuclear potential, comprised of a short-range attractive (Strong interaction) and a long-range repulsive term (Coulomb potential), is assigned to each data point. Thus, nucleons that are densely distributed in space fuse to build nuclei (clusters), whereas single point clusters are repelled (noise). The algorithm is optimized and tested in an extensive study with a series of synthetic datasets as well as the Iris data. The results show that it can robustly identify clusters even when complex configurations and noise are present. Finally, to address the initial goal, quantitative MRI data of 42 patients are analyzed employing NPC. A series of experiments with different sets of image-based features show a consistent grouping tendency: younger patients with low disease grade are recognized as cohesive clusters, while those of higher age and impairment are recognized as outliers. This allows for the definition of a reference region in a feature space associated with phenotypic data. Tracking of the individual's positions therein can disclose patients at risk and be employed for therapy evaluation.