Refine
Year of publication
- 2014 (24) (remove)
Document Type
- Doctoral Thesis (24) (remove)
Keywords
- Pestizid (2)
- Pflanzenschutzmittel (2)
- monitoring (2)
- Abdrift (1)
- Annotation (1)
- Anpassung (1)
- Araneae (1)
- Augenbewegung (1)
- Auslese (1)
- Auswahl (1)
Institute
- Fachbereich 7 (8)
- Fachbereich 5 (3)
- Institut für Informatik (3)
- Fachbereich 8 (2)
- Institut für Erziehungswissenschaft (2)
- Institute for Web Science and Technologies (2)
- Fachbereich 6 (1)
- Institut für Anglistik und Amerikanistik (1)
- Institut für Computervisualistik (1)
- Institut für Grundschulpädagogik (1)
- Institut für Integrierte Naturwissenschaften, Abt. Geographie (1)
- Institut für Kunstwissenschaft und Bildende Kunst (1)
- Institut für Pädagogik, Abteilung Schulpädagogik/Allgemeine Didaktik (1)
- Institut für Wirtschafts- und Verwaltungsinformatik (1)
The goal of this thesis is the development of methods for augmented image synthesis using 3D photo collections. 3D photo collections are representations of real scenes automatically generated from single photos and describe a scene as a set of images with known camera poses as well as a sparse point-based model of the scene geometry. The main goal is to perform a photo-realistic augmented image synthesis of real and virtual parts, where the real scene is provided as a 3D photo collection. Therefore, three main problems are addressed.
Since the photos may be represented in different device-specific RGB color spaces, a color characterization of the 3D photo collections is necessary to gain correct color information that is consistent with human perception. The proposed novel method automatically transforms all images into a common RGB color space and thereby simplifies color characterization of 3D photo collections.
As a main problem for augmented image synthesis, all environmental lighting has to be known in order to apply illumination to virtual parts that is consistent with the real portions shown in the photos. To solve this problem, two novel methods were developed to reconstruct the lighting from 3D photo collections.
In order to perform image synthesis for arbitrary views on the scene, an image-based approach was developed that generates new views in 3D photo collections making direct use of its point cloud. The novel method creates new views in real-time and allows free-navigation.
In conclusion, the proposed novel methods show that 3D photo collections are a useful representation for real scenes in Augmented Reality and they can be used to perform a realistic image synthesis of real and virtual portions.
This study examined whether positive resource change and two types of baseline stressors would influence general life satisfaction and satisfaction with health at the end of treatment in mothers attending an inpatient rehabilitation or preventive program for mothers, fathers and children. Parenting self-efficacy was included as an indicator of resources while baseline distress was operationalized in terms of parenting-related stressors and depressive symptoms, respectively. Drawing on Hobfoll- conservation of resources theory, it was hypothesized that resource change and baseline stressors would interact in their effect on satisfaction measures at the end of the preventive/rehabilitation program: resource gains were expected to influence satisfaction more strongly in patients who reported higher stress levels at baseline (as compared to patients with lower baseline distress; primary research question). Hypotheses were tested using a sample of N=1724 female prevention and rehabilitation patients by means of structural equation modeling including latent interactions. Results showed that parenting-related stressors were negatively associated with general satisfaction as well as satisfaction with health while resource gains had a small positive effect on satisfaction. There was, however, no significant interaction of parenting-related stressors and positive resource change. When depressive symptoms were included as a measure of baseline distress, they were negatively associated with both facets of satisfaction. In these models, there were only small positive associations of resource gains and satisfaction measures. Again, no interaction effects were found. Additional analyses that comprised both stressor types as predictors showed that the effects of depressive symptoms on satisfaction were more pronounced than the influence of parenting-related stressors; the effects of resource gains were small to negligible. In addition, it was examined whether problem-oriented coping strategies mediated the influence of positive resource change on general life satisfaction and satisfaction with health, respectively (secondary research question). Correlations between coping, resource change and satisfaction were low or not significant. Given this lack of direct associations between the relevant variables, no further path analytic testing was performed; the hypothesized mediating role of problem-oriented coping could thus not be confirmed. All in all, study results show that contrary to the hypotheses derived from conservation of resources theory, the effects of positive resource change and distress indicators on patients" life satisfaction and satisfaction with health were not interdependent.
By the work presented in this thesis, the CH4 emissions of the River Saar were quantified in space and time continuously and all relevant processes leading to the observed pattern were identified. The direct comparison between reservoir zones and free-flowing intermediate reaches revealed, that the reservoir zones are CH4 emission hot spots and emitted over 90% of the total CH4. On average, the reservoir zones emitted over 80 times more CH4 per square meter than the intermediate reaches between dams (0.23 vs. 19.7 mol CH4 m-2 d-1). The high emission rates measured in the reservoir zones fall into the range of emissions observed in tropical reservoirs. The main reason for this is the accumulation of thick organic rich sediments and we showed that the net sedimentation rate is an excellent proxy for estimating ebullitive emissions. Within the hot spot zones, the ebullitive flux enhanced also the diffusive surface emissions as well as the degassing emissions at dams.
To resolve the high temporal variability, we developed an autonomous instrument for continuous measurements of the ebullition rate over long periods (> 4 weeks). With this instrument we could quantify the variability and identify the relevant trigger mechanisms. At the Saar, ship-lock induces surges and ship waves were responsible for over 85% of all large ebullition events. This dataset was also used to determine the error associated with short sampling periods and we found that with sampling periods of 24 hours as used in other studies, the ebullition rates were systematically underestimated by ~50%. Measuring the temporal variability enabled us to build up a conceptual framework for estimating the temporal pattern of ebullition in other aquatic systems. With respect to the contribution of freshwater systems to the global CH4 emissions, hot spot emission sites in impounded rivers have the potential to increase the current global estimate by up to 7%.
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.
In the recent years, Software Engineering research has shown the rise of interest in the empirical studies. Such studies are often based on empirical evidence derived from corpora - collections of software artifacts. While there are established forms of carrying out empirical research (experiments, case studies, surveys, etc.), the common task of preparing the underlying collection of software artifacts is typically addressed in ad hoc manner.
In this thesis, by means of a literature survey we show how frequently software engineering research employs software corpora and using a developed classification scheme we discuss their characteristics. Addressing the lack of methodology, we suggest a method of corpus (re-)engineering and apply it to an existing collection of Java projects.
We report two extensive empirical studies, where we perform a broad and diverse range of analyses on the language for privacy preferences (P3P) and on object-oriented application programming interfaces (APIs). In both cases, we are driven by the data at hand, by the corpus itself, discovering the actual usage of the languages.
Through the increasing availability of access to the web, more and more interactions between people take place in online social networks, such as Twitter or Facebook, or sites where opinions can be exchanged. At the same time, knowledge is made openly available for many people, such as by the biggest collaborative encyclopedia Wikipedia and diverse information in Internet forums and on websites. These two kinds of networks - social networks and knowledge networks - are highly dynamic in the sense that the links that contain the important information about the relationships between people or the relations between knowledge items are frequently updated or changed. These changes follow particular structural patterns and characteristics that are far less random than expected.
The goal of this thesis is to predict three characteristic link patterns for the two network types of interest: the addition of new links, the removal of existing links and the presence of latent negative links. First, we show that the prediction of link removal is indeed a new and challenging problem. Even if the sociological literature suggests that reasons for the formation and resolution of ties are often complementary, we show that the two respective prediction problems are not. In particular, we show that the dynamics of new links and unlinks lead to the four link states of growth, decay, stability and instability. For knowledge networks we show that the prediction of link changes greatly benefits from the usage of temporal information; the timestamp of link creation and deletion events improves the prediction of future link changes. For that, we present and evaluate four temporal models that resemble different exploitation strategies. Focusing on directed social networks, we conceptualize and evaluate sociological constructs that explain the formation and dissolution of relationships between users. Measures based on information about past relationships are extremely valuable for predicting the dissolution of social ties. Hence, consistent for knowledge networks and social networks, temporal information in a network greatly improves the prediction quality. Turning again to social networks, we show that negative relationship information such as distrust or enmity can be predicted from positive known relationships in the network. This is particularly interesting in networks where users cannot label their relationships to other users as negative. For this scenario we show how latent negative relationships can be predicted.
Mathematical Modelling of GIS Tailored GUI Design with the Application of Spatial Fuzzy Logic
(2014)
This PhD thesis is situated within the framework of the Research-Group Learning and Neurosciences (ReGLaN)-Health and Logistics project. The goal of this project is the optimisation of health service delivery in the rural areas of South Africa. Cooperation takes place between ReGLaN-Health and Logistics and the South African Council for Scientific and Industrial Research (CSIR) Meraka Institute, with Prof Dr Dr Marlien Herselman of Pretoria, South Africa, as the central contact person. This thesis deals with the mathematical modelling of Geographic Information System (GIS)-tailoredrnGraphical User Interface (GUI) design with the application of spatial fuzzy logic. This thesis considers the mathematical visualisation of risk and resource maps for epidemiological issues using GIS and adaptive GUI design for an Open Source (OS) application for digital devices. The intention ofrnthis thesis is to provide spatial decision support tailored to different user groups. In order for the GUI elements to be evaluated and initialised, empirical teaching-learning-research on dealing with geomedia and GUI elements was conducted.
This study explored the question whether greenhouse gas mitigation projects in Namibia could be initiated through local economic development programmes. In particular, research was done on whether the Clean Development Mechanism (CDM) of the Kyoto Protocol could play an essential role in the promotion of such mitigation projects.
In a first step supporting and inhibiting factors (potential for mitigation projects, business and investment climate, institutions, etc.) were discussed, which have a negative or positive influence on mitigation projects. In a second step the mind-set of climate and energy experts as well as of local economic development experts and practitioners was analysed with regard to the research questions. To this end, 229 questionnaires, 28 interviews and the output of a focus group discussion with 20 participants were evaluated. Additionally, the author conducted a real life case study to investigate the practicability of initiating greenhouse gas mitigation projects through local economic development efforts. Parallel to the development of an economic development strategy in the Namibian region of Otjozondjupa, the potential for greenhouse gas mitigation projects was explored. Based on the outcome of this investigation project ideas were developed and their potential socio-economic impact was evaluated. Promising projects were then included into the development strategy.
Due to various factors such as the complexity of CDM, low greenhouse gas emissions in Namibia, the low price of emission rights and insufficient financial means it is unlikely that CDM projects can be initiated through local economic development initiatives in Namibia. However, many stakeholders consider the idea of interlinking mitigation projects and local economic development initiatives favourably as long as such projects support the broader objectives of those initiatives. This research has shown that locally initiated mitigation projects do not contribute much to employment or income generation at the local level in Namibia. Thus, national strategic objectives should be considered, such as improving access to electricity to all strata of society or becoming less dependent on electricity imports. This requires, however, that local economic development strategies also cover the energy sector, that local governments are willing and capacitated to initiate mitigation projects, that national and local public institutions work together more closely, that national and local economic framework conditions are improved so as to attract private investments, and that the experiences and interests of the relevant stakeholders are considered throughout the project development process.
Diffusion imaging captures the movement of water molecules in tissue by applying varying gradient fields in a magnetic resonance imaging (MRI)-based setting. It poses a crucial contribution to in vivo examinations of neuronal connections: The local diffusion profile enables inference of the position and orientation of fiber pathways. Diffusion imaging is a significant technique for fundamental neuroscience, in which pathways connecting cortical activation zones are examined, and for neurosurgical planning, where fiber reconstructions are considered as intervention related risk structures.
Diffusion tensor imaging (DTI) is currently applied in clinical environments in order to model the MRI signal due to its fast acquisition and reconstruction time. However, the inability of DTI to model complex intra-voxel diffusion distributions gave rise to an advanced reconstruction scheme which is known as high angular resolution diffusion imaging (HARDI). HARDI received increasing interest in neuroscience due to its potential to provide a more accurate view of pathway configurations in the human brain.
In order to fully exploit the advantages of HARDI over DTI, advanced fiber reconstructions and visualizations are required. This work presents novel approaches contributing to current research in the field of diffusion image processing and visualization. Diffusion classification, tractography, and visualizations approaches were designed to enable a meaningful exploration of neuronal connections as well as their constitution. Furthermore, an interactive neurosurgical planning tool with consideration of neuronal pathways was developed.
The research results in this work provide an enhanced and task-related insight into neuronal connections for neuroscientists as well as neurosurgeons and contribute to the implementation of HARDI in clinical environments.
Field margins are often the only remaining habitats of various wild plant species in agricultural landscapes. However, due to their proximity to agricultural fields, the vegetation of field margins can be affected by agrochemicals applied to the crop fields. The aim of this thesis was to investigate the individual and combined effects of herbicide, insecticide and fertilizer inputs on the plant community of a field margin. Therefore, a 3-year field experiment with a randomized block design including seven treatments (H: herbicide, I: insecticide, F: fertilizer, H+I, F+I, F+H and F+H+I) and one control was conducted on a low-production meadow. Each treatment was replicated 8 times in 8 m x 8 m plots with a distance of 2 m between each plot. The fertilizer rates (25 % of the field rate) and pesticide rates (30 % of the field rate) used for the plot applications were consistent with realistic average input rates (overspray + drift) in the first meter of a field margin directly adjacent to a wheat field.
The study revealed that fertilizer and herbicide misplacements in field margins are major factors that affect the natural plant communities of these habitats. In total, 20 of the 26 abundant species on the study site were significantly affected by the fertilizer and herbicide treatment. The fertilizer promoted plants with high nutrient uptake and decreased the frequencies of small species. The herbicide caused a nearly complete disappearance of three species directly after the first application, whereas sublethal effects (e.g., phytotoxic effects and reduced seed productions of up to 100 %) were observed for the other affected species. However, if field margins are exposed to repeated agrochemical applications over several years, then such sublethal effects (particularly reproduction effects) also reduce the population size of plant species significantly, as observed in this study.
Significant herbicide-fertilizer interaction effects were also detected and could not be extrapolated from individual effects. The fertilizer and herbicide effects became stronger over time, leading to shifts in plant community compositions after three years and to a 15 % lower species diversity than in the control. The insecticide significantly affected the frequencies of two plant species (1 positively and 1 negatively). The results of the experiment suggest that a continuous annual agrochemical application on the study site would cause further plant community shifts and would likely lead to the disappearance of certain affected plants. A clear trend of increasing grass dominance at the expense of flowering herbs was detected. This finding corresponds well with monitoring data from field margins near the study site.
Although herbicide risk assessment aims to protect non-target plants in off-field habitats from adverse effects, reproduction effects and combined effects are currently not considered. Furthermore, no regulations for fertilizer applications next to field margins exist and thus, fertilizer misplacements in field margins are likely to occur and to interact with herbicide effects.
Adaptations of the current risk assessment, a development of risk mitigation measures (e.g., in-field buffers) for the application of herbicides and fertilizers, and general management measures for field margins are needed to restore and conserve plant diversity in field margins in agricultural landscapes.
Diese Studie leistet einen Beitrag zur Weiterentwicklung der Nachrichtenwerttheorie. Hierzu wird das Konzept der Fotonachrichtenfaktoren zunächst theoretisch aufgearbeitet und seine Plausibilität anhand verschiedener Teilbereiche der Journalismusforschung dargelegt. Darauf aufbauend wird es im Rahmen einer Inhaltsanalyse eingesetzt, die die Prüfung des erweiterten Modells der Nachrichtenwerttheorie ermöglicht. Dieses Modell geht von einem gemeinsamen Einfluss der textbasierten Nachrichtenfaktoren und der Fotonachrichtenfaktoren auf den Nachrichtenwert bzw. die journalistische Beachtung eines Beitrags aus. Ergänzend werden die Fotonachrichtenfaktoren zur Bildung von Fototypen herangezogen. Die Analysen werden am Beispiel der Kriegs- und Krisenberichterstattung deutscher Tageszeitungen durchgeführt. Dieser Themenbereich des Journalismus genießt viel Aufmerksamkeit sowohl bei Medienrezipienten als auch bei Kommunikationswissenschaftlern. Die Ergebnisse der Studie unterstützen die Grundidee der Fotonachrichtenfaktoren: Sie können, ähnlich wie die textbasierten Nachrichtenfaktoren, zu übergeordneten Bündeln zusammengefasst werden. Diese Bündel zeigen Einflüsse auf die formale Gestaltung der Fotos und können gemeinsam mit den Textnachrichtenfaktoren teilweise die journalistische Beachtung des Artikels erklären. Darüber hinaus ist die gewählte Operationalisierung der Fotonachrichtenfaktoren geeignet, Fototypen bzw. zentrale Motive der visuellen Kriegsberichterstattung deutscher Tageszeitungen zu identifizieren. Die Ergebnisse dieser Studie liefern Ausgangspunkte für weitere Fragestellungen zum Zusammenwirken der Text- und Fotonachrichtenfaktoren, der Anwendung der Zwei-Komponenten-Theorie und der Wirkung von typischen Bildmotiven.
In an attempt to put pronunciation training back in the limelight and to engage in a fruitful discussion on pronunciation- equal importance with other language skills, the current research project focuses on pronunciation. Specifically, one area that causes problems for L1-German speakers is concentrated on, namely the central approximant. German learners should not despair too much as even many native-English children stumble with /r/. For some, acquiring this sound takes a long time and even some professional help (see Secord 2007: 7-8). One problem that these children have is the substitution of /r/ with /w/. "Children with developmental speech substitution," write Raphael et al., "often produce the more easily articulated (…) /w/ for /r/" (2011: 119). Most native-English children require a long time to acquire this sound. According to Bleile, /r/ belongs to the last stage (i.e. stage 4) of a child- language development (see Bleile 2004: 106). This final stage can range from five years of age to adolescence. One of the greatest difficulties, for instance, occurs with postvocalic /r/ (or "[r] colored schwa") and consonant clusters with /r/ (see Bleile 2004: 106).
It is important at this juncture to direct attention to the organization of the current research project and how it addresses the /r/-problem. At the start of the current study, an appropriate and concise definition of the /r/-sound is proposed. While it can be easily stated that the letter <r> is the 18th letter in the English alphabet, it is much more arduous to define what is meant by the /r/-sound. In chapter 2, a number of definitions of /r/ are explored. Section 2.1 presents an articulatory description and section 2.2 offers an acoustic description. A theoretical definition is provided in section 2.3. By exploring the distinguishing features of the /r/-sound, it can be distinguished from other sounds. From these three perspectives, a more encompassing view of the complicated English /r/-sound can be obtained.
In chapter 3, three hypotheses are presented. The first two are the main hypotheses for the current study. There is also a third (minor) hypothesis and it is an offshoot from some of the research findings from the first hypothesis. In section 3.1.1, hypothesis 1 (H1) investigates the frequency of the problem that L1-German speakers have when speaking the American English /r/-sound aloud when it is postvocalic and in a weakened syllable. Hypothesis 2 (H2) is put forward in section 3.1.2. It explores the frequency of /w/-substitution for /r/ when the latter sound is prevocalic in a stressed syllable. For the final and third hypothesis (H3), one specific occassion where L1 interference occurs (i.e. /ɐ/ for /ə/) is researched in section 3.1.3.
Some organizational details concerning the study are also included in the third chapter. The general set-up of the study is provided in section 3.2.1. Information pertaining to both the disqualified participants of the study as well as the 50 test subjects who were selected for the study is available in section 3.2.1.2. Details on technical equipment and software programs are disclosed in sections 3.2.1.1 and 3.2.1.3. The prescribed reading texts that the test subjects needed to read aloud are presented in the subsections of section 3.2.2.
For a better understanding of the acoustic properties of the prevocalic /r/-sound, five expressions from the reading text are selected. In section 3.3, the expressions are also articulated by the author (a native speaker of a variety of English from North America) in four different scenarios. The scenarios try to illustrate a limited number of possible articulatory variations. The author utters the expressions in four different ways: 1.) with lip spreading, 2.) in isolation, 3.) in connected speech and 4.) with /w/-substitution for /r/. These articulations serve as a template for understanding how the acoustic features of /r/ changes. Some of these changes are noticeable in the articulations of the test subjects (presented in chapter 4).
In chapter 4, the data collected for the three hypotheses are presented. For H1, section 4.1 and its subsections document the formant readings of a number of English expressions which contain the postvocalic /r/. In this analysis, it is important to detect where the central approximant is elided. In cases of elision, the samples have their schwa-like sounds plotted in a vowel quadrilateral, so that their locations can be detected. In section 4.2 and its subsections, data observations for H2 are discussed. From these data, the extent of /w/-substitution for /r/ when it is prevocalic in a stressed syllable can be observed. In section 4.3, an analysis of the /ɐ/-sound from three German expressions as they are spoken by the 50 test subjects is presented for H3. Together with pertinent data (i.e. formant readings) from H1, a comparison is made between the locations of /ə/ and /ɐ/. This section concludes with a plotting of all samples of these two vowels in a vowel chart.
In the final chapter of the current study, the main findings of the three hypotheses are summarized. Chapter 5 also provides some insights into a few of the gaps or limitations in the present research. Future research endeavors are proposed in order to suggest possible avenues to advance the findings reported on in the fourth chapter. In addition, several teaching suggestions and pronunciation exercises for implementing the /r/-sound in class are presented.
Lastly, there is an appendix section that consists of 14 smaller appendices (i.e Volume 2). The spectrograms and formant readings for each expression spoken by each of the 50 test subjects are documented in Appendix A to M. In Appendix N, some charts, tables, diagrams, etc. that are discussed throughout the following chapters are made available. Copies of the original questionnaire that were filled out by the test subjects are also in Appendix N.
The availability of digital cameras and the possibility to take photos at no cost lead to an increasing amount of digital photos online and on private computers. The pure amount of data makes approaches that support users in the administration of the photo necessary. As the automatic understanding of photo content is still an unsolved task, metadata is needed for supporting administrative tasks like search or photo work such as the generation of photo books. Meta-information textually describes the depicted scene or consists of information on how good or interesting a photo is.
In this thesis, an approach for creating meta-information without additional effort for the user is investigated. Eye tracking data is used to measure the human visual attention. This attention is analyzed with the objective of information creation in the form of metadata. The gaze paths of users working with photos are recorded, for example, while they are searching for photos or while they are just viewing photo collections.
Eye tracking hardware is developing fast within the last years. Because of falling prices for sensor hardware such as cameras and more competition on the eye tracker market, the prices are falling, and the usability is increasing. It can be assumed that eye tracking technology can soon be used in everyday devices such as laptops or mobile phones. The exploitation of data, recorded in the background while the user is performing daily tasks with photos, has great potential to generate information without additional effort for the users.
The first part of this work deals with the labeling of image region by means of gaze data for describing the depicted scenes in detail. Labeling takes place by assigning object names to specific photo regions. In total, three experiments were conducted for investigating the quality of these assignments in different contexts. In the first experiment, users decided whether a given object can be seen on a photo by pressing a button. In the second study, participants searched for specific photos in an image search application. In the third experiment, gaze data was collected from users playing a game with the task to classify photos regarding given categories. The results of the experiments showed that gaze-based region labeling outperforms baseline approaches in various contexts. In the second part, most important photos in a collection of photos are identified by means of visual attention for the creation of individual photo selections. Users freely viewed photos of a collection without any specific instruction on what to fixate, while their gaze paths were recorded. By comparing gaze-based and baseline photo selections to manually created selections, the worth of eye tracking data in the identification of important photos is shown. In the analysis of the data, the characteristics of gaze data has to be considered, for example, inaccurate and ambiguous data. The aggregation of gaze data, collected from several users, is one suggested approach for dealing with this kind of data.
The results of the performed experiments show the value of gaze data as source of information. It allows to benefit from human abilities where algorithms still have problems to perform satisfyingly.
German politicians have identified a need for greater citizen involvement in decision-making than in the past, as confirmed by a recent German parliamentarians study ("DEUPAS"). As in other forms of social interactions, the Internet provides significant potential to serve as the digital interface between citizens and decision-makers: in the recent past, dedicated electronic participation ("e-participation") platforms (e.g. dedicated websites) have been provided by politicians and governments in an attempt to gather citizens" feedback and comment on a particular issue or subject. Some of these have been successful, but a large proportion of them are grossly under-used " often only small numbers of citizens use them. Over the same time period, enthusiasm of Society for social networks has increased and is now commonplace. Many citizens use social networks such as Facebook and Twitter for all kinds of purposes, and in some cases to discuss political issues.
Social networks are therefore obviously attractive to politicians " from local government to federal agencies, politicians have integrated social media into their daily work. However, there is a significant challenge regarding the usefulness of social networks. The problem is the continuous increase in digital information: social networks contain vast amounts of information, and it is impossible for a human to manually filter the relevant information from the irrelevant (so-called "information overload"). Even using the search tools provided by social networks, it is still a huge task for a human to determine meanings and themes from the multitude of search results. New technologies and concepts have been proposed to provide summaries of masses of information through lexical analysis of social media messages, and therefore they promise an easy and quick overview of the information.
This thesis examines the relevance of these analyses" results, for the use in everyday political life, with the emphasis on the social networks Facebook and Twitter as data sources. Here we make use of the WeGov Toolbox and its analysis components that were developed during the EU project WeGov. The assessment has been performed in consultation with actual policy-makers from different levels of German government: policy-makers from the German Federal Parliament, the State Parliament North Rhine-Westphalia, the State Chancellery of the Saarland and the cities of Cologne and Kempten all took part in the study. Our method was to execute the analyses on data collected from Facebook and Twitter, and present the results to the policy-makers, who would then evaluate them using a mixture of qualitative methods.
The responses of the participants have provided us with some useful conclusions:
1) None of the participants believe that e-participation is possible in this way. But participants confirm that "citizen-friendliness" can be supported by this approach.
2) The most likely users for the summarisation tools are those who have experience with social networks, but are not "power users". The reason being is that "power users" already knew the relevant information provided by analysis tools. But without any experiences for social networks it is hard to interpret the analysis results the right way.
3) The evaluation has considered geographical aspects, and related this to e.g. a politician- constituency as a local area of social networks. Comparing the rural to the urban areas, it is shown that the amount of relevant political information in the rural areas is low. While the proportion of publicly available information in urban areas is relatively high, the proportion in the rural areas is much lower.
The findings that result from the engagement with policy-makers will be systematically surveyed and validated within this thesis.