1 - 6 of 6
rss atomLink til resultatlisten
Permanent link
Referera
Referensformat
  • apa
  • harvard1
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
  • Disputas: 2019-12-16 09:00 J1516, Karlskrona
    Klotins, Eriks
    Blekinge Tekniska Högskola, Fakulteten för datavetenskaper, Institutionen för programvaruteknik.
    Software Engineering in Start-up Companies2019Doktoravhandling, med artikler (Annet vitenskapelig)
    Abstract [en]

    Start-up companies have emerged as suppliers of innovation and software-intensive products. Small teams, lack of legacy products, experimental nature, and absence of any organizational processes enable start-ups to develop and market new products and services quickly. However, most start-ups fail before delivering any value. Start-up failures can be attributed to market factors, shortcomings in business models, lack of motivation, or self-destruction, among other reasons. However, inadequacies in product engineering precede any market or business-related challenges and could be a significant contributing factor to start-up failures. At the same time, state-of-the-art software engineering (SE) practices are often neglected by start-ups as inadequate. At the beginning of this work, SE in start-ups had attracted very little attention from researchers. Thus, there was no coherent view of SE state-of-practice in start-ups and no starting point for a focused investigation.

     

    In this thesis, we explore how start-ups practice SE, what specific SE challenges should be addressed, and what new SE practices are needed to support the engineering of innovative software-intensive products and services.

     

    A substantial part of this work is exploratory and aimed to explore SE state-of-practice in start-ups. Our initial findings suggest that start-ups overlook the best SE practices. Teams of a few people working on relatively experimental and straightforward software see no upside of following the best practices. However, late start-ups face substantial challenges as their teams grow, and products become more complex. The key difficulties concern installing adequate SE practices supporting collaboration, coordination of work, and management of accumulated technical debt. To support the evolution of engineering practices in start-ups, we propose the start-up progression model outlining engineering goals, common challenges, and useful practices with regards to the start-up life-cycle phases. Further findings suggest inadequate support for market-driven requirements engineering (MDRE). Specifically, on how to aggregate needs and wishes of a large and loosely defined set of stakeholders who may not be able to articulate their needs and expectations. To address this challenge, we propose a method for the identification and prioritization of data sources and stakeholders in MDRE. Analyzing SE context in start-ups and other organizations developing innovative and market-driven products, we have found many similarities. While start-ups have challenges, they do not appear to be unique. Thus, most start-up challenges can be addressed by transferring the best practices from other engineering contexts.

     

    We conclude that there is a little need for start-up specific engineering practices. Best software engineering practices are relevant to address challenges in start-ups. The key engineering challenge in start-ups is the management of the evolution of SE practices to match the growing complexity of the product and the organization. Our work also highlights the need for better MDRE practices to support new market-driven product development in both start-ups and other types of organizations. 

  • Disputas: 2019-12-17 14:00 J1640, Karlskrona
    Eivazzadeh, Shahryar
    Blekinge Tekniska Högskola, Fakulteten för teknikvetenskaper, Institutionen för hälsa.
    Evaluating Success Factors of Health Information Systems2019Doktoravhandling, med artikler (Annet vitenskapelig)
    Abstract [en]

    Health information systems are our technological response to the growing demand for health care. However, their success in their mission can be challenging due to the complexity of evaluating technological interventions in health care. In the series of studies compiled in this dissertation, we looked at the evaluation of these systems. We focused on the evaluation of factors that lead to success, where success is indicated by user satisfaction and can be induced by both intervention-specific and individual-specific factors.

    Study 1 developed a method, called UVON, to elicit and organise the user-demanded qualities in the outcomes of the health information system intervention. Through the application of the UVON method in the FI-STAR project, an EU project which developed and deployed seven e-health applications in seven member countries, ten categories of quality and their subcategories were identified. These qualities formed two questionnaires, specific to the patient and health professional users. Through the questionnaires, the patients and health-professionals users evaluated and graded both the occurrence of those demanded qualities in the project outcomes and their general satisfaction.

    Study 2 analysed the survey results to find out which of those ten qualities have the highest impact on satisfaction or can predict it better. Two partial least squares structural equation modelling (PLS-SEM) models were constructed, for the patient and health professionals, based on the Unified eValuation using ONtology (UVON) and survey outputs. The models showed that effectiveness is an important quality in creating satisfaction for both user groups. Besides, affordability for the health professionals and efficiency plus safety for the patients were the most influential. A satisfaction index is also introduced for simple and fast inferring of the changes in the outcome qualities.

    Study 5 recruited outputs and learnings from studies 1 and 2 to design a system that partially automates the process of evaluating success factors in health information systems, making it continuous and real-time, and replacing hard-to-run surveys with automatically captured indicators and analytics.

    Study 3 focused on individual-specific factors in using health information systems, particularly the technophilia personality trait. A short six-items instrument, called TechPH, was designed to measure technophilia in users, tuned for older users. The study recruited empirical data from the Swedish National Study on Aging and Care (SNAC) project. Two factors, labelled techAnxiety and techEnthusiams, are identified by the factor analysis method. A TechPH score was introduced as a scalar measurement of technophilia.

    Study 4 elicited and discussed the ethical challenges of evaluating and researching health information systems. Both a scoping review and a novel systematic postulation approach were recruited to identify twenty ethical challenges. The identified ethical challenges were discussed and mapped into a three-dimensional space of evaluation stages, demanded qualities, and major involving entities (stakeholder and artefacts), which fosters further postulation of ethical challenges.

  • Disputas: 2019-12-19 10:00 Rio Grande, KarlshamnGulbrandsen, Elisabeth
    The New Production of Politics: Between the No Longer and the Not Yet2019Doktoravhandling, med artikler (Annet vitenskapelig)
    Abstract [en]

    Indications that the global environmental and inequality crises are intimately linked to our western ways of living, challenge the self-understanding of participants in the modern research-complex. As researchers we not only observe, unveil, analyse and solve problems "out there". Our knowledge-producing activities are (re)productive forces whose effects are not contained by the walls of any "ivory tower". As researchers we do not have a standpoint outside of a research-dependent culture. We are implicated in it. How do we convert this implication into resources for transformative movements in science and society?

    The main objective motivating the texts presented has been to explore conditions for developing responsible technoscientific cultures - in and beyond - the academy. The linearity as well as the division of labour suggested by the "technology push" and "society pull" policy models are heavily criticized for ignoring the complexity and dynamics that emerge partly as a consequence of the success and pervasiveness of science and technology in late modernity. Science and society have both become transgressive invading each other's domains, and science policy questions are enhanced into political questions. A third, more interactive policy model is emerging figured in transdiscursive terms like 'strategic science', 'innovation system', 'post-normal science', 'technoscience', 'mode 2' and 'agora'.

    The more specific objective has been to situate research processes as "triple loop" learning processes and to figure both 'research quality' and 'politics' in innovative ways that help responsible technoscientific cultures emerge. Resources from European traditions of "action learning" and "action research" as well as the recent U.S. trend of "technoscience as culture" are employed as frameworks for the analysis. Conditions for responsible innovation are explored through trying transformations or "participant provocations" at the University of Oslo (1986-1994) and at the Research Council of Norway (1998-2017). These experiments are documented in published articles and function as "original communications" to the thesis.

  • Disputas: 2019-12-20 09:30 J1650 (Campus Gräsvik), Karlskrona
    Carlsson, Liesel
    Blekinge Tekniska Högskola, Fakulteten för teknikvetenskaper, Institutionen för strategisk hållbar utveckling.
    Inviting Community into the Development of Globally Sustainable Food Systems2019Doktoravhandling, med artikler (Annet vitenskapelig)
    Abstract [en]

    Food systems and human diets contribute to unsustainable socioecological conditions, which in turn negatively affect human health. These driver-impact relationships span multiple scales, prompting international governance bodies, nations, and communities alike to grapple with solutions for a better food future. Collaborative action across scales and sectors is necessary; however, how communities can align contributions with efforts at broader scales is unclear.

    The aim of this research is to develop theoretical and procedural supports for community engagement in globally sustainable food systems (SFS), and to provide concrete results relevant to one case community.

    The community of nutrition and dietetics professionals was chosen as the case community given its history of engagement with SFS, its integration throughout food system sectors, and because dietary shifts have significant potential to contribute to SFS. Furthermore, the researcher’s position as a member of this community supported the case study work.

    The research uses transdisciplinary methods guided by the Framework for Strategic Sustainable Development (FSSD) and Community Development theory. The FSSD provides a concrete definition of sustainability and includes methodological supports for co-creation of sustainability transitions. Community Development theory supports participatory approaches and welcomes different knowledge cultures in such co-creation. The Delphi Inquiry method was used to facilitate data collection and community engagement. For measurement-specific elements of the research, causal loop diagrams (CLD) informed by the Cultural Adaptation Template (CAT) theory were used, and Critical Dietetics was used as a framework for dietetics-specific analysis.

    High level insights include that: (i) participatory and multidimensional approaches are important to facilitate community engagement in SFS development; (ii) objective parametres for defining sustainability are critical to guide concerted action and can provide an innovation space that invites creative and diverse solutions within; (iii) systems thinking and related tools help simplify the complexity of food systems without disregarding broader context, and support assessment in the absence of all data. Specifically in relation to the case community explored, insights include that, (i) integrating an SFS lens into existing roles and activities is important, because dietitians already work across sectors and scales, making them well positioned to contribute in diverse ways; (ii) a shared language based on transdisciplinary understandings of SFS is required; (iii) engaging in activities that facilitate SFS knowledge development within the profession, prior to integrating it into roles and activities, is an important first step; (iv) collaborative and reflexive approaches to continued knowledge development and practice are important, such that in the end sustainability becomes integrated into a cultural way of thinking about food.  

    Based on these insights, this dissertation outlines a procedure for collaborative community work for globally SFS. The procedure is adaptable to various community settings. The dissertation also provides specific guidance for how dietitians could utilise their strategic positions throughout food systems to contribute to SFS development.

  • Disputas: 2020-01-24 13:00 C413A, Karlskrona
    Andersén, Jimmie
    Blekinge Tekniska Högskola, Fakulteten för teknikvetenskaper, Institutionen för fysisk planering.
    Ett politiskt bygge: - översiktsplaneringens innehållsmässiga utveckling2020Doktoravhandling, monografi (Annet vitenskapelig)
    Abstract [en]

    Debate often describes the evolution of spatial planning as a process of neo-liberalisation. However, this evolution has not been studied from the perspective of planning in general in Sweden. Whether comprehensive planning is also undergoing neo-liberalisation is considered here based on these overarching research questions: 

    -          To what extent has spatial planning undergone neo-liberalisation?

    -          How can the evolution of comprehensive planning be understood as a neo-liberalisation process?

    Answering these questions requires an understanding of the content characterising this planning, and of what a neo-liberalisation process entails. To achieve such an understanding, this dissertation examines how planning in general can be understood as both a political control system and a market-based land use regime.

    The discourse of comprehensive planning was studied by analysing 17 comprehensive plans adopted in five municipalities over the 1989–2014 period.

    Generally, these plans follow a few main evolutionary paths. Municipal comprehensive planning has transitioned from being instrumental, encountering few problems, to becoming increasingly abstract and all-encompassing. Comprehensive planning increasingly concerns how growth is to be achieved locally. This change is occurring in both form and content. Change in form refers to the fact that comprehensive planning has been encompassing more and more subject areas over time, whereas change in content refers to the increasingly market-based content of such planning.

    The research question regarding the extent to which spatial planning is undergoing neo-liberalisation is answered by the obvious trend towards neo-liberalisation. This trend is not pervasive, as countervailing tendencies are also present. Spatial planning is thus becoming internally contradictory, with earlier values persisting, albeit often in interaction with new ones.

    The question of how comprehensive planning can be understood as a neo-liberalisation process has a three-part answer:

    -          As content in comprehensive plans, neo-liberalisation can be understood as a focus on market rationality. 

    -          As form in comprehensive plans, neo-liberalisation can be understood as policy production with content that assumes an ever more neo-liberal orientation. Comprehensive planning during the studied period has come to include more and more municipal areas of activity.

    -          From a neo-liberal perspective, neo-liberalisation is understood as economic utility maximisation.

    Planning is becoming increasingly neo-liberalised, but not neo-liberal. That this has occurred as a result of increasingly comprehensive policy formulations is somewhat surprising. We are faced with the politicisation of planning in general, which is increasingly informed by market-based policy. This trend represents a comprehensive policy shift towards more neo-liberal planning content, accompanied by more policies. More principles are appearing, addressing how planning and development are to proceed compared with the content of pre-existing plans. The neo-liberalisation process is thus occurring via means that are contradictory from a neo-liberal perspective.

  • Disputas: 2020-01-31 09:59 Karlskrona
    García-Martín, Eva
    Blekinge Tekniska Högskola, Fakulteten för datavetenskaper, Institutionen för datavetenskap.
    Energy Efficiency in Machine Learning: Approaches to Sustainable Data Stream Mining2020Doktoravhandling, med artikler (Annet vitenskapelig)
    Abstract [en]

    Energy efficiency in machine learning explores how to build machine learning algorithms and models with low computational and power requirements. Although energy consumption is starting to gain interest in the field of machine learning, still the majority of solutions focus on obtaining the highest predictive accuracy, without a clear focus on sustainability.

    This thesis explores green machine learning, which builds on green computing and computer architecture to design sustainable and energy efficient machine learning algorithms. In particular, we investigate how to design machine learning algorithms that automatically learn from streaming data in an energy efficient manner.

    We first illustrate how energy can be measured in the context of machine learning, in the form of a literature review and a procedure to create theoretical energy models. We use this knowledge to analyze the energy footprint of Hoeffding trees, presenting an energy model that maps the number of computations and memory accesses to the main functionalities of the algorithm. We also analyze the hardware events correlated to the execution of the algorithm, their functions and their hyper parameters.

    The final contribution of the thesis is showcased by two novel extensions of Hoeffding tree algorithms, the Hoeffding tree with nmin adaptation and the Green Accelerated Hoeffding Tree. These solutions are able to reduce their energy consumption by twenty and thirty percent, with minimal effect on accuracy. This is achieved by setting an individual splitting criteria for each branch of the decision tree, spending more energy on the fast growing branches and saving energy on the rest.

    This thesis shows the importance of evaluating energy consumption when designing machine learning algorithms, proving that we can design more energy efficient algorithms and still achieve competitive accuracy results.