http://aran.library.nuigalway.ie/:80 The ARAN digital repository system captures, stores, indexes, preserves, and distributes digital research material. Fri, 19 Mar 2021 17:12:03 GMT 2021-03-19T17:12:03Z http://hdl.handle.net/10379/16610 Efficient path query approaches over distributed linked data Mehmood, Qaiser In the past decade, graphs have become the defacto data model in some of the popular application domains such as bioinformatics, social, and road networks. One of the most used variants of the graph model is the labeled graph where vertices and edges are given names. In the con- text of Semantic Web technologies, this graph representation is called Linked Data, expressed using RDF (Resource Description Format), where entities are given their names with Uniform Resource Identifiers (URIs) which uniquely identify each entity. A plethora of such information, corresponding to various domains, exists over the Web and is exposed as knowledge bases. Some of the prominent examples of such knowledge bases or datasets are YAGO, DBPedia and Bio2RDF. Things or entities described within these datasets can be interconnected within a single graph or these connections may lead to different datasets. This large collection of interlinked and distributed data over the web has created new chal- lenges in the context of data management, data integration and knowledge discovery. One of the key challenges is querying and analysing relevant information from such large-scale knowledge bases. For instance, in the context of graph analysis, navigational queries are at the basis of core tasks and processes. Several approaches have been proposed to perform path queries, with some supporting declarative queries, while others are based on imperative constructs. More- over, some approaches work in a distributed manner while others in a central way. For instance, SPARQL1.1, which is a declarative query language, introduced a feature called Property Paths (PPs). Using PPs, a user can check the existence of paths between two entities within a single graph. However, this variant is not very expressive, does not capture some of the important properties of linked data, and has performance issues when dealing with large-scale data. In order to exploit the full potential of linked data, the standard declarative query language of SPARQL needs (1) to be equipped with expressive features to capture some important graph analytical tasks and (2) to perform better even for large-scale graphs. We propose an extension to SPARQL1.1 property paths which we believe complements the current version of PPs with analytical features in a more expressive and robust manner. The algorithm proposed in our extension is a relatively straightforward solution but performs efficiently compared to tailored solutions in the literature. Our solution allows users to compute k shortest paths matching a property expression between two nodes. While the constant growth of linked data on the web in various domains has led to a challenging environment for data processing, it has also justified the need for novel approaches to handle and analyse such big data. Ecosystems such as Hadoop or Pregel have become defacto standards for such tasks and are capable of navigating the path queries in a distributed manner. However, these systems are developed for customised data distribution within a controlled environment, and in the context of Linked Data such systems are not considered purely heterogeneous. In our second major contribution in this thesis, considering the limitations and management efforts to handle homogeneous systems, we propose different solutions which we believe are easy to adopt and scalable. They do not require complicated management settings when the data is either customized in a controlled environment or even if the data is accessible publicly in a heterogeneous environment. Our first implementation towards distributed path query processing is FedS, which we pro- pose as a P2P solution. However, our solution in contrast with the core P2P architecture is of a hybrid type. Generally, queries in P2P networks are blindly forwarded from node to node, and also a loose guarantee of resource discovery in P2P network can make it possible not to find a resource node although it does exist. We introduced the concept of source selection within the P2P network, which reduces the number of query requests by selecting the relevant source nodes. Our second implementation for distributed architecture is QPPDS, which is an index-based path query engine with the concept of finding paths within a heterogeneous environment. We proposed a source selection mechanism based on a pre-computed index. We also provide a dataset, exposed as a SPARQL endpoint, for the paths calculated by QPPDS. This dataset contains the statistical and provenance information for those paths. We provide a comprehensive evaluation of this approach on real-world life sciences data with different execution strategies. Our third implementation is a cache-based and index-free distributed path query engine called DpcLD, which computes the paths distributed within a homogeneous environment. We provide extensive evaluations on both synthetic and real-world data. We evaluate DpcLD not only against the distributed approaches but also do a comparison with the centralized approach and show trade-offs in different contexts. In this thesis, we proposed different approaches to address the pathfinding task, especially in a distributed environment. From the individual contributions described above, the overall conclusion of this work is that graph management and distribution have a strong impact on the efficiency and accuracy of pathfinding, and therefore that different solutions are required depending on their characteristics. Thu, 18 Mar 2021 00:00:00 GMT http://hdl.handle.net/10379/16610 2021-03-18T00:00:00Z http://hdl.handle.net/10379/16609 Transitional justice in Africa: Traditional and modern approaches to addressing genocide, war crimes and crimes against humanity De Ycaza, Carla The term transitional justice generally refers to a range of judicial and nonjudicial mechanisms that have been utilized by countries in order to address massive human rights abuses. Such mechanisms can include prosecutions, truth commissions, amnesties, reparations and institutional reforms, as well as traditional methods of conflict resolution. It can be argued that the development of transitional justice as a field is most evident on the African continent. Africa provides a living laboratory for transitional justice with a wide range of mechanisms implemented throughout the continent, and has set precedents in both establishing truth commissions and pursuing criminal prosecutions. Africa has also provided unique examples of the interaction between international and localized mechanisms of transitional justice. Each situation must be evaluated within its own context in order to determine appropriate responses to mass atrocity; however, there are trends in what approaches to transitional justice have been more utilized and ultimately more effective in addressing the goals of transitional justice than others. Although these trends cannot be labeled as an all-encompassing African approach to transitional justice, identifying these patterns could help international and local actors support effective transitional justice mechanisms on the ground, rather than implementing justice processes from above. To date, there has been no comprehensive analysis of the question of whether there is an African approach to transitional justice, and if there is, what this approach entails. Through a comparative case study analysis of transitional justice mechanisms in Africa, this research infers broader universal knowledge claims from particular context-specific sources in order to analyze what are the main features of transitional justice in Africa in response to violations of international law including genocide, war crimes and crimes against humanity. Tue, 01 Nov 2016 00:00:00 GMT http://hdl.handle.net/10379/16609 2016-11-01T00:00:00Z http://hdl.handle.net/10379/16608 Design, synthesis, and public understanding and involvement in randomised trials Finucane, Elaine Introduction The initial aim of my PhD was to assess if amniotic membrane sweeping, a common intervention in maternity care, is a safe and effective way of preventing a formal induction of labour1 in pregnant women at or near term2. However, with the arrival of a novel coronavirus (SARS-CoV-2) and the immediate and profound changes it brought to clinical practice and the conduct of clinical research, the intended pathway of my PhD research altered significantly. Therefore, my thesis comprises two sections. Section one focuses on membrane sweeping to prevent a formal induction of labour. Section two of the thesis focuses on The People’s Trial. The aim of The People’s Trial is to support the public understanding and knowledge of randomised trials, to understand why they matter and be better equipped to think critically about health claims by becoming involved in each step of the clinical trial process. The People’s Trial also potentially supports researchers in learning how best to involve and engage the public in trials. Methods This thesis includes four papers. The first paper (Chapter 2 published), a Cochrane systematic review and meta-analysis was undertaken to assess the effects and safety of amniotic membrane sweeping for induction of labour in women at or near term. The findings of this review directly informed The MILO Study, presented in paper two (Chapter 3 published). The MILO Study is a Feasibility study protocol of a pragmatic, randomised controlled pilot trial, to evaluate the effectiveness (including optimal timing and frequency) of membrane sweeping to prevent post-term pregnancy. Paper three (Chapter 4, submitted for review), describes the process of The People’s Trial, an online parallel group, randomised controlled trial, including aims, design, conduct and dissemination. Paper four (Chapter 5, submitted for review), reports the design, conduct and findings of the randomised controlled trial designed by the public, which we called The Reading Trial. Results The Cochrane systematic review and meta-analysis found that when comparing membrane sweeping with no treatment/sham, women randomised to membrane sweeping may be more likely to experience a spontaneous onset of labour and less likely to experience an induction of labour; however, overall, the evidence was of low certainty. The review identified the need for further robust research to assess the optimal gestation to receive a membrane sweep and whether having more than one sweep would be beneficial. It further highlighted the need to explore women’s views of membrane sweeping. The MILO study, a feasibility study, includes a pilot randomised trial, a health economic analysis, a qualitative study and a Study Within A Trial (SWAT). The MILO study was due to commence recruitment in March 2020, with ethical approval, study documentation, site procedures, clinicians and research midwives in position to support the study conduct. The MILO Study is now due to commence recruitment in February 2021, dependent on clinical circumstances and COVID-19. The remainder of the PhD focuses on, The People’s Trial, a novel, online randomised trial designed by the public for the public. Over 3000 members of the public, from 72 countries, participated in The People’s Trial, engaging in all aspects of the trial design, from choosing the trial question, to trial conduct, analysis and dissemination. We report the processes of The People’s Trial in seven phases, mimicking the steps of a randomised trial, In December 2019, 991 participants took part in a trial designed by the public, called The Reading Trial. The trial aimed to answer a question identified and prioritised by the public ‘Does reading a book in bed make a difference to sleep in comparison to not reading a book in bed?’ The Reading Trial found that, 56/369 (42%) of people in the intervention group felt their sleep improved, compared to 112/405 (28%) of those in the control group, a difference of 14%. Conclusion My PhD supports the development of four papers, which individually and collectively provide an original contribution to knowledge. The outputs from this body of work include: A multi-site feasibility study, informed by a Cochrane systematic review, which is designed and ready to commence recruitment in 2021. The findings of the MILO study, including the views of women and clinicians, will inform the optimal design of a future definitive randomised trial to examine the effectiveness including optimal timing and frequency of membrane sweeping to prevent post-term pregnancy. The People’s Trial, a novel, online trial that has successfully involved over 3000 members of the public in the design, conduct, and dissemination of a randomised trial. This demonstrates public appetite to engage with, and learn about randomised trials, to understand why they matter, and to be better equipped to think critically about health claims. Tue, 16 Mar 2021 00:00:00 GMT http://hdl.handle.net/10379/16608 2021-03-16T00:00:00Z http://hdl.handle.net/10379/16607 Challenging social norms to recraft practices: A Living Lab approach to reducing household energy use in eight European countries Sahakian, Marlyne; Rau, Henrike; Grealis, Eoin; Godin, Laurence; Wallenborn, Grégoire; Backhaus, Julia; Friis, Freja; Genus, Audley T.; Goggins, Gary; Heaslip, Eimear; Heiskanen, Eva; Iskandarova, Marfuga; Jensen, Charlotte Louise; Laakso, Senja; Musch, Annika-Katrin; Scholl, Christian; Vadovics, Edina; Vadovics, Kristof; Vasseur, Véronique; Fahy, Frances ENERGISE is the first large-scale European effort to reduce household energy use through a change initiative that adopted a ‘living lab’ approach informed by social practice theory. Two challenges were introduced to 306 households in eight countries: to lower indoor temperatures and to reduce laundry cycles. This contribution demonstrates the usefulness of a practice-centered design that takes habits and routines as an entry point for understanding how different ‘elements of practices’ can be re-crafted. We discuss how a participatory ‘living lab’ approach that explicitly encouraged deliberation and reflexivity served to sharpen attention on practices as central to change. We discuss how ‘doing laundry’ and ‘keeping warm’, as very different types of practices, responded to the change initiative. For laundry, tangible changes in material arrangements, news skills and sensory competencies, and shifts in what is seen as ‘normal’ proved to be central to reducing wash cycles, including wearing clothes more often, airing them out, using smell to gauge cleanliness, or keeping dirty clothes out of sight. Warming people rather than spaces through added layers and activities, and related shifts in norms around thermal comfort, emerged as crucial steps towards lowering indoor temperatures. Average changes in reported temperatures and wash cycles indicate that reductions are possible, without an emphasis on individuals or technologies as central to change. We end with a discussion on the implications of our approach for energy sufficiency thinking and practice, emphasizing the merits of taking the complexity of everyday life seriously when designing change initiatives. Wed, 10 Feb 2021 00:00:00 GMT http://hdl.handle.net/10379/16607 2021-02-10T00:00:00Z