1.
Chatterjee, Bapi, 1982
(författare)
Efficient Implementation of Concurrent Data Structures on Multi-core and Many-core Architectures
2015
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
Synchronization of concurrent threads is the central problem in order to design efficient concurrent data-structures. The compute systems widely available in market are increasingly becoming heterogeneous involving multi-core Central Processing Units (CPUs) and many-core Graphics Processing Units (GPUs). This thesis contributes to the research of efficient synchronization in concurrent data-structures in more than one way. It is divided into two parts. In the first part, a novel design of a Set Abstract Data Type (ADT) based on an efficient lock-free Binary Search Tree (BST) with improved amortized bounds of the time complexity of set operations - Add, Remove and Contains, is presented. In the second part, a comprehensive evaluation of concurrent Queue implementations on multi-core CPUs as well as many-core GPUs are presented.Efficient Lock-free BST -To the best of our knowledge, the lock-free BST presented in this thesis is the first to achieve an amortized complexity of O(H(n)+c) for all Set operations where H(n) is the height of a BST on n nodes and c is the contention measure. Also, the presented lock-free algorithm of BST comes with an improved disjoint-access-parallelism compared to the previously existing concurrent BST algorithms. This algorithm uses single-word compare-and-swap (CAS) primitives. The presented algorithm is linearizable. We implemented the algorithm in Java and it shows good scalability.Evaluation of concurrent data-structures - We have evaluated the performance of a number of concurrent FIFO Queue algorithms on multi-core CPUs and many-core GPUs. We studied the portability of existing design of concurrent Queues from CPUs to GPUs which are inherently designed for SIMD programs. We observed that in general concurrent queues offer them to efficient implementation on GPUs with faster cache memory and better performance support for atomic synchronization primitives such as CAS. To the best of our knowledge, this is the first attempt to evaluate a concurrent data-structure on GPUs.
2.
Man, Yemao, 1987
(författare)
Human-Machine Interface Considerations for Design and Testing in Distributed Sociotechnical Systems
2015
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
The increasing concerns for safety and environmental sustainability create demands on the development of future maritime transportation strategies. One way to meet these demands is the concept of autonomous unmanned vessels for intercontinental voyages. As automation is being introduced onboard and watch keeping operations being migrated to the shore, there is a risk introducing new human factor issues among the various stakeholder groups and add to the complexity of the actors’ roles. This licentiate was based on the context of an EU research project MUNIN (Maritime Unmanned Ship through Intelligence in Networks) about remote monitoring and controlling autonomous unmanned ships where the bridge and engine control room were moved from the ship to a land based control station.Human Machine Interface, as a mediating artefact in the complex system to bridge automation/engine control is of importance for situation awareness, reliability, efficiency, effectiveness, resilience and safety. The purpose of the thesis is to achieve a comprehensive understanding of the complexity of Human Machine Interface in a distributed complex system by exploring the experiences of the human agents during the designing and testing phases of a designed for purpose Human Machine Interface. The results reveal prominent human factor issues related to situation awareness and automation bias within such a complex distributed sociotechnical system, which sheds light on the design considerations of Human Machine Interface. Loss of presence can lead to critical perceptual bottlenecks which could negatively impact upon the operators; the organizational factors also greatly shape individual and team performance. It indicates that the contextual factors in the distributed sociotechnical system must be accommodated by the interface design through a holistic systemic approach. The Human Machine Interface shall not only support data visualization, but also the process and context in which data are utilized and understood for consensus decision-making.
3.
Granåsen, Magdalena, 1978-
(författare)
Exploring C2 Capability and Effectiveness in Challenging Situations : Interorganizational Crisis Management, Military Operations and Cyber Defence
2019
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
Modern societies are affected by various threats and hazards, including natural disasters, cyber-attacks, extreme weather events and inter-state conflicts. Managing these challenging situations requires immediate actions, suspension of ordinary procedures, decision making under uncertainty and coordinated action. In other words, challenging situations put high demands on the command and control (C2) capability. To strengthen the capability of C2, it is vital to identify the prerequisites for effective coordination and direction within the domain of interest. This thesis explores C2 capability and effectiveness in three domains: interorganizational crisis management, military command and control, and cyber defence operations. The thesis aims to answer three research questions: (1) What constitutes C2 capability? (2) What constitutes C2 effectiveness? and (3) How can C2 effectiveness be assessed? The work was carried out as two case studies and one systematic literature review. The main contributions of the thesis are the identification of perspectives of C2 capability in challenging situations and an overview of approaches to C2 effectiveness assessment. Based on the results of the three studies, six recurring perspectives of capability in the domains studied were identified: interaction (collaboration), direction and coordination, relationships, situation awareness, resilience and preparedness. In the domains there are differences concerning which perspectives that are most emphasized in order obtain C2 capability. C2 effectiveness is defined as the extent to which a C2 system is successful in achieving its intended result. The thesis discusses the interconnectedness of performance and effectiveness measures, and concludes that there is not a united view on the difference between measures of effectiveness and measures of performance. Different approaches to effectiveness assessment were identified, where assessment may be conducted based on one specific issue, in relation to a defined goal for a C2 function or using a more exploratory approach.
4.
Löfström, Tuve
(författare)
Utilizing Diversity and Performance Measures for Ensemble Creation
2009
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
An ensemble is a composite model, aggregating multiple base models into one predictive model. An ensemble prediction, consequently, is a function of all included base models. Both theory and a wealth of empirical studies have established that ensembles are generally more accurate than single predictive models. The main motivation for using ensembles is the fact that combining several models will eliminate uncorrelated base classifier errors. This reasoning, however, requires the base classifiers to commit their errors on different instances – clearly there is no point in combining identical models. Informally, the key term diversity means that the base classifiers commit their errors independently of each other. The problem addressed in this thesis is how to maximize ensemble performance by analyzing how diversity can be utilized when creating ensembles. A series of studies, addressing different facets of the question, is presented. The results show that ensemble accuracy and the diversity measure difficulty are the two individually best measures to use as optimization criterion when selecting ensemble members. However, the results further suggest that combinations of several measures are most often better as optimization criteria than single measures. A novel method to find a useful combination of measures is proposed in the end. Furthermore, the results show that it is very difficult to estimate predictive performance on unseen data based on results achieved with available data. Finally, it is also shown that implicit diversity achieved by varied ANN architecture or by using resampling of features is beneficial for ensemble performance.
5.
Crusoe, Jonathan, 1991-
(författare)
Why is it so challenging to cultivate open government data? : Understanding impediments from an ecosystem perspective
2019
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
Introduction: This compilation licentiate thesis focuses on open government data (OGD). The thesis is based on three papers. OGD is a system that is organized when publishers collect and share data with users, who can unrestrictedly reuse the data. In my research, I have explored why it can be challenging to cultivate OGD. Cultivation is human activities that change, encourage, or guide human organizations towards a higher purpose by changing, introducing, managing, or removing conditions. Here, the higher purpose is OGD to realize believed benefits. Thus, OGD cultivation is an attempt to stimulate actors into organizing as OGD.Problem and Purpose: OGD is believed to lead to several benefits. However, the worldwide OGD movement has slowed down, and researchers have noted a lack of use. Publishers and users are experiencing a set of different impediments that are challenging to solve. In previous research, there is a need for more knowledge about what can impede the OGD organization, cause non-valuable organizing, or even collapse the organization. At the same time, there is a lack of knowledge about how impediments shape the organization of OGD. This gap can make it hard to solve and overcome the impediments experienced by publishers and users. The sought-after knowledge can bring some understanding of the current situation of OGD. In this research, I have viewed the organization of OGD as an ecosystem. The purpose of this thesis is to draw lessons about why it can be challenging to cultivate OGD ecosystems by understanding OGD impediments from an ecosystem perspective.Research Design: I set out to explore OGD through qualitative research from 2016 to 2018. My research started with a pilot case study that led to three studies. The studies are each reported in a paper and the papers form the base of this thesis. The first paper aims to stimulate the conceptually oriented discussion about actors’ roles in OGD by developing a framework that was tested on a Swedish public agency. The second paper has the purpose of expanding the scope surrounding impediments and was based in a review and systematization of previous research about OGD impediments. The third paper presents an exploration of impediments experienced by publishers, users, and cultivators in the Swedish national OGD ecosystem to identify faults. From the three papers, lessons were drawn in turn and together, that are presented in this thesis.Findings: Cultivators when cultivating OGD ecosystems are facing towering challenges. The following three main challenges are identified in this thesis: (1) to cultivate a system that can manage stability by itself without constant involvement, (2) to cultivate a system that is capable of evolving towards a “greater good” by itself, and (3) to have an up-to-date precise vocabulary for a self-evolving system that enables inter-subjective understand for coordinating problem-solving.Contribution: The theoretical contribution of this thesis is that OGD ecosystems can be viewed as a public utility. Moreover, I recommend that researchers approach the organizing of OGD as the cultivation of evolution, rather than the construction of a structure; to consider the stability of the system in growth, value, and participation; and to be cautious with how they label and describe OGD actors. For actors that are cultivating OGD, I recommend that they guide the OGD actors to help them organize; view OGD cultivation as the management of evolution (growth) towards a purpose; and view cultivation as a collaborative effort where they can supply ideas, technologies, practices, and expertise.
6.
Alshareef, Hanaa, 1985
(författare)
A Collaborative Access Control Framework for Online Social Networks
2019
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
Online social networks (OSNs) are one of the most popular web-based services for people to communicate and share information with each other. With all their benefits, OSNs might raise serious problems in what concerns users' privacy. One privacy risk is caused by accessing and sharing co-owned data items, i.e., when a user posts a data item that involves other users, some users' privacy may be disclosed, since users generally have different privacy preferences regarding who can access and share their data. Another risk is caused by the privacy settings offered by OSNs that do not, in general, allow fine-grained enforcement, especially in cases where posted data items concern other users. We discuss and give examples of these issues, in order to illustrate their impacts on current OSNs' privacy protection mechanisms. We propose a collaborative access control framework to deal with such privacy issues. Basically, in our framework, the decision whether a user can access or share a co-owned data item is based on the aggregated opinion of all users involved. Our solution is based on the sensitivity level of users with respect to the concerned data item, the trust among users, the types of controllers (those who are concerned in making the collaborative decision) and the types of accessors (those who are identified to access a given data item or not). In order to observe how varying some of the parameters mentioned above influence the outcome of the permitting/denying decision of the proposed solution, we provide an evaluation of our framework. We also present a proof-of-concept implementation of our approach in the open source OSN Diaspora.
7.
Angulo, Julio, 1980-
(författare)
Usable privacy for digital transactions : Exploring the usability aspects of three privacy enhancing mechanisms
2012
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
The amount of personal identifiable information that people distribute over different online services has grown rapidly and considerably over the last decades. This has led to increased probabilities for identity theft, profiling and linkability attacks, which can in turn not only result in a threat to people’s personal dignity, finances, and many other aspects of their lives, but also to societies in general. Methods and tools for securing people’s online activities and protecting their privacy on the Internet, so called Privacy Enhancing Technologies (PETs), are being designed and developed. However, these technologies are often seen by ordinary users as complicated and disruptive of their primary tasks. In this licentiate thesis, I investigate the usability aspects of three main privacy and security enhancing mechanisms. These mechanisms have the goal of helping and encouraging users to protect their privacy on the Internet as they engage in some of the steps necessary to complete a digital transaction. The three mechanisms, which have been investigated within the scope of different research projects, comprise of (1) graphical visualizations of service providers’ privacy policies and user-friendly management and matching of users’ privacy preferences “on the fly”, (2) methods for helping users create appropriate mental models of the data minimization property of anonymous credentials, and (3) employing touch-screen biometrics as a method to authenticate users into mobile devices and verify their identities during a digital transaction. Results from these investigations suggest that these mechanisms can make digital transactions privacy-friendly and secure while at the same time delivering convenience and usability for ordinary users.
8.
Jurkeviciute, Monika, 1985
(författare)
Planning of a holistic summative eHealth evaluation: The interplay between standards and reality
2018
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
Summative evaluation assesses outcomes, produces evidence, and advances knowledge of eHealth implementations in health care provisions. Therefore, its quality is essential to reap benefits from the results generated by evaluation studies. Standardization is considered as among the solutions to improve quality of evaluations as it creates methodological uniformity and enhances research credibility. However, standards are used insufficiently in eHealth evaluation practice. The purpose of this thesis is to study the use of standards in eHealth evaluation planning practice, in order to provide insight to support knowledge development. This thesis takes a stance in research concerning eHealth evaluation planning with regards to the standardization, translation, and collaboration. This thesis is built on a systematic literature review assessing the extent to which eHealth evaluation frameworks are used in summative eHealth evaluations and a single case study of eHealth evaluation of a project concerning eHealth implementation within several health care contexts. It focuses on the evaluation planning process, assesses adequacy of eHealth evaluation planning guidelines to practice, and identifies reasons that hinder the use of and adherence to standards. The thesis confirms that eHealth evaluation frameworks are not used in the empirical eHealth evaluations. In contrast, the frameworks and the evaluation planning guidelines are found to be adequate and beneficial to practice. The reasons hindering the use of standards and affecting adherence to them are the insufficient evaluator’s experience and resources using a standard, evaluator’s unawareness of a standard, inadequacy of a standard to address a target population or a disease, non-existence of a validated version of a standard in a particular location, and a lack of fit between a standard and a scope of the evaluation. A model is developed, suggesting that standards can be viewed as objects translated in a specific context and influenced by collaborative activities. The thesis suggests that adherence to standards in eHealth evaluation practice could be seen as a range that is caused by trade-offs made when standards are translated based on reality.
9.
Tuma, Katja, 1991
(författare)
Towards Efficiency and Quality Assurance in Threat Analysis of Software Systems
2018
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
Context: Security threats have been a growing concern in many organizations. Organizations developing software products strive to plan for security as soon as possible to mitigate such potential threats. In the design phase of the software development life-cycle, teams of experts routinely analyze the system architecture and design to nd potential security threats. Objective: The goal of this research is to improve on the performance of existing threat analysis techniques and support practitioners with automation and tool support. To understand the inner-workings of existing threat analysis methodologies we also conduct a systematic literature review examining 26 methodologies in detail. Our industrial partners conrm that existing techniques are labor intensive and do not provide quality guarantees about their outcomes. Method: We conducted empirical studies for building an in-depth understanding of existing techniques (Systematic Literature Review (SLR), controlled experiments). Further we rely on empirical case studies for ongoing validation of an attempted technique performance improvement. Findings: We have found that using a novel risk-rst approach can help reduce the labor while producing the same level of outcome quality in a shorter period of time. Further, we suggest that the key for a successful application of this approach is two fold. First, widening the analysis scope to end-to-end scenarios guides the analyst to focus on important assets. Second, appropriate model abstractions are required to manage the cognitive load of the human analysts. We have also found that reasoning about security in a formal setting requires extending the existing notations with security semantics. Further, minimal model extensions for doing so include security contracts for system nodes handling sensitive information. In such a setting, the analysis can be automated and can to some extent provide completeness guarantees. Future work: In the future, we plan to further study the analysis completeness guarantees. In particular, we plan to improve on the analysis automation and investigate complementary techniques for analysis completeness (namely informal pattern based techniques). We also plan to work on the disconnect between the planned and implemented security.
10.
Cakici, Baki, 1984-
(författare)
Disease surveillance systems
2011
Licentiatavhandling (övrigt vetenskapligt/konstnärligt) abstract
Recent advances in information and communication technologies have made the development and operation of complex disease surveillance systems technically feasible, and many systems have been proposed to interpret diverse data sources for health-related signals. Implementing these systems for daily use and efficiently interpreting their output, however, remains a technical challenge. This thesis presents a method for understanding disease surveillance systems structurally, examines four existing systems, and discusses the implications of developing such systems. The discussion is followed by two papers. The first paper describes the design of a national outbreak detection system for daily disease surveillance. It is currently in use at the Swedish Institute for Communicable Disease Control. The source code has been licenced under GNU v3 and is freely available. The second paper discusses methodological issues in computational epidemiology, and presents the lessons learned from a software development project in which a spatially explicit micro-meso-macro model for the entire Swedish population was built based on registry data.
