Publikationer från Stockholms universitet
Not a member yet
31237 research outputs found
Sort by
Empirical analysis on the effect of multi-pivot partitioning on the parallel quicksort algorithm
No full textMulti-pivot quicksort implementation has seen a rise in popularity since the first successful implementation in 2009. There have been plenty of research showing multi-pivot quicksort in its sequential form. Currently, there is a lack of research looking at parallel solutions for the multi-pivot algorithms. This thesis, explores, with experimentation, how pivot count might affect the parallel quicksort algorithm. The result given from this experimentation shows a positive relationship between the two variables. So the conclusion reached is that increases in pivot count increases the execution time of the parallel quicksort algorithm
Upplevt socialt stöd från Generativ AI på arbetsplatsen : Utvecklande av en skala för mätning
No full textThis study investigated how employees perceive social support from Generative AI (GenAI) in the workplace. Semi-structured interviews were conducted with thirteen (n=13) individuals from various industries and professional roles. The data were analyzed using deductive thematic analysis, based on the Job Demands-Resources (JD-R) model and established theory on social support, categorized as emotional, informational, instrumental, and appraisal support. The participants perceived GenAI, in its current state, to be primarily a resource offering instrumental and informational social support through task automation, process streamlining, and rapid access to information. Some participants also experienced indirect emotional support, as GenAI created a sense of security and enthusiasm. Appraisal support was perceived to a lesser extent. Instead, concerns were raised that feedback generated by GenAI could be perceived as impersonal. The study highlighted the importance of balancing GenAI support with human interaction to create a positive and productive work environment. Based on the interview results and insights from existing literature, 25 potential scale items were generated for a quantitative questionnaire intended to measure perceived social support from GenAI tools in the workplace. These scale items were reviewed by an expert panel for feedback on their clarity and relevance. This feedback was subsequently incorporated into the final design of the scale items. The study also identified a need for further research, as the rapid technological development and adoption of these tools have the potential to significantly impact workplaces in both the short and long term
The Influence of AI Literacy on User Preferences for Explainable AI in Recommender Systems
No full textThe increasing reliance on AI-driven recommender systems, presents a challenge due to the lack of transparency. This challenge raises societal concerns such as misinformation and lack of critical thinking in decision-making, which can lead to users’ over relying on AI. Explainable AI (XAI) seeks to address this issue by offering clear explanations, thereby improving understanding and trust. However, there is a limited amount of research on how users with different levels of AI knowledge interact and prefer these explanations. This study investigated how users’ AI literacy levels influence their preferences for different explanation formats and complexities in recommender systems. Using a quantitative method with a questionnaire of 104 participants, the study assessed AI literacy using the Artificial Intelligence Literacy Scale (AILS) and gathered preferences for various explanation styles. The Chi-Square Test of Independence revealed no statistically significant association between AI literacy and the participants’ most preferred explanation style. A Linear Mixed Model (LMM) analysis showed that while AI literacy did not significantly influence format explanation preferences, it had a significant interaction with explanation complexity. Specifically, low AI literacy participants significantly preferred simple explanations over detailed ones. The other interactions AI literacy, format, and complexity did not result in significant results. These findings highlight the need for personalized XAI explanations tailored to users’ AI literacy levels, particularly concerning the complexity. This research contributes to the XAI research field by highlighting the end user’s perspective as well as contributing to the development of a more accessible recommender systems that encourage decision-making with AI
Securing Agile Development: Analysing the Intersection of Cybersecurity and Agile Methodologies - A Systematic Literature Review
No full textIntroduction This thesis explores how cybersecurity is added to Agile software development. Agile methods are used because they give flexibility and help teams deliver software quickly. However, these benefits can make it difficult to include strong security practices. Security usually needs clear andwell-organised steps, which are not always easy to follow in Agile workflows. Research Question The main research question of this thesis is: “What are the key challenges faced by Agile teams when integrating cybersecurity practices into the software development lifecycle?” Method This thesis uses a systematic literature review (SLR) method and follows the PRISMA framework to find, study, and summarise research that already exists. The review included 46 studies in the final synthesis. Results The review found four main challenges: cultural, process, technical and regulatory. Cultural issues were the most common, including low security awareness and unclear roles. Process and technical problems came from mismatched workflows and weak tools. Regulatory demands were hard to meet in fast Agile cycles. Discussion Adding cybersecurity to Agile is still challenging. Most challenges arise from team-related issues, such as unclear responsibilities, limited collaboration, and tools that are not well suited to Agile practices. Cultural gaps show that security is not a shared priority. Solutions such as DevSecOps exist, but are not widely used. These findings challenge the assumption that cultural change alone is sufficient, instead suggesting that successful integration of cybersecurity in Agile workflows requires structural rethinking of sprint governance, tool alignment, and incentive systems to extend existing DevSecOps models
Sign language interpreters' considerations in ethical dilemmas
No full textTolkar ställs i sitt yrke inför situationer som kräver etiska överväganden och snabba beslut. Trots etikens centrala betydelse i tolkyrket är tolkars egna resonemang kring etiska dilemman ett område som ännu är relativt outforskat. Denna studie undersöker och analyserar tolkars förhållningssätt i etiskt utmanande situationer. Studiens metod utgörs av en enkätundersökning med vinjetter, där deltagarna besvarat frågor med både fasta svarsalternativ och möjlighet att skriva fritextkommentarer. De fasta svarsalternativen, det vill säga den kvantitativa delen av svaren har redovisats genom deskriptiv statistik där respektive svar har relaterats till normativ etik utifrån plikt- respektive konsekvensetik. Den kvalitativa delen som utgörs av fritextsvar, har resulterat i tre centrala teman per vinjett. Studien visar att tolkar reflekterar på olika sätt kring yrkesetiska principer, och att deras förhållningssätt påverkas av faktorer såsom uppdragets allvarlighetsgrad, kontexten och relationen till parterna.Interpreters are faced with situations in their profession that require ethical considerations and quick decisions. Despite the central importance of ethics in the interpreting profession, interpreters' own reasoning regarding ethical dilemmas is an area that remains relatively unexplored. This study examines and analyzes interpreters' approaches in ethically challenging situations. The study's method consists of a questionnaire survey with vignettes, in which participants answered questions with both fixed response options and the opportunity to write free-text comments. The fixed response options, i.e., the quantitative part of the responses, have been reported using descriptive statistics, where each response has been related to normative ethics based on duty-based and consequence-based ethics. The qualitative part, consisting of free-text responses, has resulted in three central themes per vignette. The study shows that interpreters reflect in different ways on professional ethical principles, and that their approach is influenced by factors such as the seriousness of the assignment, the context, and the relationship with the parties involved
Studenters behov av generativ AI i högre utbildning
No full textDen snabba utvecklingen av generativ AI har haft en inverkan på högre utbildning och skapat nya möjligheter för personligt anpassat och tillgängligt lärande. Denna studie undersöker hur generativ AI kan fungera som ett pedagogiskt stöd utifrån studenters faktiska behov och preferenser. Genom en kvalitativ forskningsdesign genomfördes sju semistrukturerade intervjuer med studenter inom data- och systemvetenskap vid svenska universitet. En tematisk analys resulterade i tre huvudteman: Förbättrad interaktion och kommunikation med AI, AI för strukturellt stöd och navigering i kursinnehåll och förutsättningar för meningsfull AI-användning. Studenterna uttryckte en önskan om AI-verktyg som är personligt anpassade, integrerade i lärplattformar och tillgängliga. Studien drar slutsatsen att användarcentrerad design och pedagogisk integration är avgörande för den framtida utvecklingen av AI-baserade lärandestöd. Begränsningar i studien inkluderar ett litet och homogent urval. Framtida forskning bör undersöka generativ AI:s långsiktiga effekter som lärandestöd samt olika studentgruppers behov i bredare utbildningskontexter.The rapid development of generative AI has significantly impacted higher education, offering new possibilities for personalized and accessible learning. This study explores how generative AI can serve as a pedagogical support tool based on students actual needs and preferences. Using a qualitative research design, seven semi-structured interviews were conducted with students in computer and systems sciences programs at Swedish universities. Through thematic analysis three main themes emerged: improved interaction and communication with AI, AI as structural support in course navigation and the conditions for meaningful AI usage in education. Students expressed a desire for AI tools that are personalized, integrated with learning platforms and accessible at all times. The study concludes that user-centered design and pedagogical integration are essential for future development of AI-based learning tools. The limitations include a small and homogeneous sample. Future research should investigate long-term effects of generative AI for learning aid and diverse student needs in broader educational contexts
A Just Transition from the Soil Up : Beyond BECCS: Agriculture and the Role of Biochar in Sweden’s Climate Strategy
No full textSweden’s commitment to achieving net-zero greenhouse gas emissions by 2045 has led to significant investment in carbon dioxide removal (CDR) technologies, primarily focusing on large-scale urban-based solutions like Bioenergy with Carbon Capture and Storage (BECCS). However, this top-down approach risks overlooking the potential and needs of rural actors. This thesis investigates the role of biochar, a promising, multifunctional CDR method with agricultural co-benefits, as a decentralized alternative in Sweden’s green transition. Through qualitative interviews with farmers currently producing and applying biochar, and guided by the theoretical framework of Just Energy Transition, the study explores the economic, technical, and social barriers they face. The findings reveal that energy justice issues, typically explored in Global North-South contexts, also manifest within Sweden, where rural producers face exclusion from policy support, certification markets, and decision-making processes. These insights suggest that a truly inclusive climate transition requires greater recognition of informal knowledge, context-specific challenges, and equitable policy development. The study concludes that successfully achieving a just transition in Sweden requires the inclusion of agricultural actors in the national CDR strategy, supported by policies that recognize their ambition, challenges, and contributions
Vad beror ojämlikheten i skolan egentligen på?
No full textI samarbete med Carlos Rojas, vid Specialpedagogiska Institutionen på Stockholms Universitet har vi i satsningen, En jämlik chans att klara skolan, undersökt sambandet mellan elevers gymnasiebehörighet och att växa upp med ekonomiska svårigheter. Statisitiken är hämtad från Statistiska Centralbyrån och analysen är Carlos personliga reflektion.Innehåller nya statistiska uppgifter rörande gymnsiebehörighet vid låg inkomstnivå och låg ekonomisk standard 2023.En jämlik chans att klara skolan – hur kan civilsamhället bidra
Social Water Management : Initiating Social Learning in a Swedish Local Context
No full textThis thesis explores whether initiating social learning can deepen local perceptions and knowledge regarding water-related issues in Lindholmen, a village north of Stockholm, Sweden, and how this can support community-based adaptation strategies for flood management. Social learning, a theory that goes beyond stakeholder participation by focusing on two-way dialogue and iterative engagement, helps individuals develop deeper understanding of climate challenges. Questions are raised about perceptions and knowledge, deepening its understanding, how municipal planning can be supported, and the challenges of implementing social learning locally. This qualitative research combines two surveys with a workshop. Data highlights how proximity to lakes, personal experience with floods, and interest in local water issues shape diverse perspectives among residents, while also showing that deepening of knowledge and collective awareness can be fostered as challenges such as limited demographic diversity and the absence of a neutral facilitator complicate the process. Nevertheless, early signs of shared understanding and a collective call for action emerged, pointing to the potential of social learning as an iterative, ongoing approach to support flood adaptation at local levels. This thesis contributes to the field of environmental planning and management by offering empirical insights into applying social learning in a Swedish local context.
From plot to Place : Individual meaning and personal value of Stockholm’s allotment gardens from a renters’ perspective
No full textWith increased demand for urban green spaces in densifying cities, the interest in urban gardens is resurrected. However, social aspects such as a renters’ ecosystem service perspective have been neglected. By focusing on five allotment gardens (AGs) in Stockholm, Sweden, consideration is put on long-established urban gardens. To explore the individual meaning and personal value of Stockholm’s AGs from the renters’ perspective the concept of Place is employed under the self-appointed name feeling of Place. The analysis seeks to depict the construction, experience, and expression of feeling of Place of Stockholm’s allotment gardeners. The data was collected through semi-structured interviews with additional observations and revealed AGs’ role for providing purpose through gardening as recreational practice creating (re)connection with nature. Crucial precondition are continuity and the behaviour in the gardens guided by the norms in the AG social sphere. Memories with the culture of gardening influence the experience, as well as social interaction, resulting in the (visible) expression of the AGs as home. Concluding, the gardens are centres of meaning which create identity and purpose. Future research and planning should increasingly consider a bottom-up perspective to understand the importance of allotment gardens as meaningful urban green spaces for their renters.