Publikationer från Stockholms universitet
Not a member yet
31237 research outputs found
Sort by
Students’ Attitudes Towards Encryption Backdoors A small-scale study among university students in Stockholm
No full textIntroduction With the rapid growth of computing and networking technology, the amount of data transferred has increased, as well as the interest in analyzing it. Advancements in cryptography have enabled secure end-to-end communication, which in turn has inspired legislative proposals advocating for encryption backdoors. This development has sparked debates concerning the balance between personal privacy and national security. Students are an integral part of society, therefore their attitudes play an important role in shaping government decisions related to encryption and privacy. Research Question What knowledge and attitudes do students have about encryption backdoors, and how are they correlated? Method This study adopts a quantitative approach and a survey research strategy is employed, enabling direct collection of students’ opinions on encryption backdoors. Data is gathered through a web-based questionnaire consisting of close-ended questions. This is a suitable method because it facilitates the collection of straight-forward information that can be quickly and easily obtained via the Internet. Open-source software is used to analyze the data through descriptive and inferential statistics. Results Survey results from 84 students indicate that most respondents had prior knowledge of end-to-end encryption, though self-rated knowledge was generally moderate. Men and IT students reported significantly higher knowledge levels than women and non-IT students. Attitudinal findings show widespread concern about the privacy implications of end-to-end encryption backdoors and potential government misuse, alongside conditional acceptance of certain use cases and strong support for legislative regulation. No significant attitude differences were found between IT and non-IT students, while men expressed stronger privacy-related concerns than women. Correlation analyses reveal that higher knowledge of end-to-end encryption is associated with more negative attitudes towards backdoors, particularly regarding privacy and government surveillance. Discussion The findings suggest that increased knowledge of end-to-end encryption is associated with more critical attitudes towards encryption backdoors, particularly regarding privacy risks. While students largely oppose backdoors, their conditional acceptance indicates a nuanced perspective reflecting the perceived trade-off between privacy and security. The study highlights the importance of public knowledge and informed debate in discussions surrounding encryption regulation and government surveillance. Despite limitations related to sample size and sampling method, the results contribute to understanding how knowledge shapes attitudes towards encryption backdoors within a student population
Avsaknad av samsyn kring tolkning och tolkens roll : Otydliga förväntningar – ett dilemma i yrkesutövningen
No full textDuring assignments interpreters are frequently asked to carry out additional tasks other than interpreting, which places them in ethical dilemmas. The aim of this qualitative study is firstly to identify and describe the types of tasks in-volved from the perspective of practicing interpreters. Sec-ondly, based on the types of tasks, the aim is to reflect on what views of interpreting and the interpreter’s role among interpreting service users this reflects. By highlighting these experiences, the study seeks to shed light on interpreters’ working conditions and to spark discussion about the prac-tice and conditions of interpreting. The study is based on a digital survey (n=FGH) targeting professional interpreters in Sweden and their experiences of this issue. The open-ended responses were analyzed using thematic analysis. The results reveal a wide variety of tasks beyond interpreting and the co-existence of parallel views of interpreting and the inter-preter’s role. The lack of shared understanding may lead to unclear and contradicting expectations that may complicate not only the interpreters’ working conditions but also nega-tively affect the interpreter-mediated encounter as such. Keywords:public sector interpreting, interpreter’s role, expectations, professional practice Nyckelord:tolkning i offentlig sektor, tolkens roll, förväntningar, yrkes-utövning
Object Recognition with LiDAR for Navigation
No full textFast object detection models like YOLO can be used for real-time detection and identification of objects. This may aid those who are Blind or Visually Impaired (BVI), who often struggle to detect objects in their path, especially indoors. When developing tools for BVI persons, smartphone apps are appropriate due to accessibility and affordability. Being small computers, smartphones can run object detection models, but their cameras cannot gather distance information. LiDAR can, though, and is included in recent iPhone models. To ensure that BVI persons can navigate safely indoors, robust navigation systems are needed. Object detection models and LiDAR in smartphones may be a foundation for such systems, making it important to evaluate these technologies. We ask: How feasible are the technologies of object detection machine learning models, in combination with LiDAR, for use in smartphones to assist BVI persons with navigation tasks? To answer this, we developed an app (using ARKit, CoreML, YOLO11 n/s/m), and performed an experiment to answer the following hypothesis: “A combination of LiDAR with a YOLO model can achieve: H1. a latency of feedback to users in the range of 200ms to 300ms, H2. energy usage below <10% battery decrease per hour, H3. a precision and recall above 90%.”. From the experiment data was collected from the application and used to calculate key performance parameters and values. Recall values were sufficiently close to 1 for s and m models, but not n. All models got sufficiently high precision scores (>90%). In the short term the latencies were in an acceptable range for all models tested. Long term, latency would spike into non ideal ranges (400ms and above) and could possibly be solved by further optimization of the application. However, all models showed too high battery usage (>20%/h), so further work in reducing battery usage by optimizing the application is needed
From Conflict to Safety and Well-being : The work of the Student Health Team to prevent physical violence between students in special needs schools
No full textTowards a Sustainable Modeling Tool for MAL
No full textThe Meta Attack Language (MAL) provides a framework for creating domain-specific languages tailored to cybersecurity modeling and analysis. However, the discontinuation of graphical editors can disrupt workflows and hinder adoption, necessitating a systematic approach to finding suitable and sustainable replacements. This paper employs Design Science Research to identify, evaluate, and select a suitable graphical modeling tool for the MAL ecosystem. We elicited requirements from five stakeholders familiar with MAL and developed a comprehensive requirements specification artifact. This artifact was then demonstrated by systematically evaluating a set of 26 candidate tools, filtering them based on mandatory criteria. The evaluation led to the identification of ADOxx as the most suitable candidate, balancing MAL’s specific meta-modeling needs with crucial sustainable factors
Power, Work, and Workplace Domination : A Republican Case for Workplace Democracy
No full textThis thesis examines whether the republican ideal of freedom as non-domination can be secured in contemporary capitalist workplaces and which institutional arrangements are required to achieve it. By drawing on republican political theory, particularly the work of Philip Pettit and his later control-based revision of freedom as non-domination, this thesis critically examines commonly suggested strategies for securing freedom as non-domination at work, such as exit-oriented strategies, workplace constitutionalism, trade unions, and the introduction of a universal basic income. While these strategies offer different answers to workplace domination, this thesis argues that none of the proposed strategies can secure the individualized, efficacious, and unconditioned control that Pettit’s control-based definition of freedom as non-domination requires. Consequently, this thesis argues that only workplace democracy can secure freedom as non-domination. It therefore concludes that workplace democracy is not merely one strategy amongst others, but the only strategy capable of securing freedom as non-domination at work
Necessary or Optional? Gender and Coalition Strategies in French Municipal Elections
No full textThis paper studies whether the gender of political leaders affects coalition strategies in electoral settings. I use French municipal elections, where institutional rules determine if coalitions are necessary or optional to enter the second round. I compare strategic choices of women and men list leaders, controlling for electoral strength, political orientation and local characteristics. The results show that women leaders are more likely to form coalitions when merging is required, but tend to avoid coalitions when they are optional. These differences have implications for the characteristics of the coalitions formed, as well as for access to seats on municipal councils
Multi-Objective Reinforcement Learning for Energy–Sensing Trade-offs in Smart IoT Systems
No full textIntroduction Efficient sensing in Internet of Things (IoT) networks often conflicts with limited battery capacity. Balancing sensing quality with energy use is a central challenge. This thesis investigates how a reinforcement learning approach can support this balance and enable more adaptive IoT systems. Research Question The main research question is: How can a tabular Multi-Objective Reinforcement Learning approach be used to effectively balance energy efficiency and sensing quality in dynamic IoT networks? Method The study follows the Design Science Research method and develops an artifact called TQ-IoT, a tabular Multi-Objective Reinforcement Learning (MORL) algorithm for IoT optimization. The research strategy is simulation, where an IoT environment is modeled with battery levels, event likelihood, and sensing outcomes. Data for evaluation are collected from repeated simulation runs and analyzed using descriptive analysis. Results The results show that TQ-IoT effectively balances energy consumption and sensing quality by adapting its actions depending on battery state and event likelihood. Energy-focused policies prioritize sleep actions and extend battery lifetime, while quality-focused policies increase sensing and transmission to improve detection performance at the cost of higher energy use. Intermediate preference settings give Pareto-efficient policies that achieve better sensing quality per unit of energy than baseline strategies. Discussion The findings confirm that a tabular MORL approach with Chebyshev scalarization is a viable framework for adaptive decision-making in energy-constrained IoT devices. By optimizing for both energy use and sensing quality, TQ-IoT enables flexible trade-offs that static or single-objective methods cannot achieve. However, the simplified simulation environment and tabular formulation limit generalizability. Future research should explore real-world deployments, adapt the model to support continuous state spaces, and provide mechanisms to handle dynamically changing preferences
Idyll och samhällsklass i Titta, Madicken, Det snöar! : Samspelet mellan bild och text i bilderboken
No full textThis thesis addresses a gap in research on the ideological dimensions of picture books by analysing Titta, Madicken, det snöar!, one of Astrid Lindgren’s more socially and ideologically conscious stories, raising questions about how these dimensions are expressed in a book for young children. The study explores how Ilon Wikland’s images interact with Lindgren’s text to convey and challenge notions of idyll, depict social class, and show the relationship between idyll and social class. Using the concept of iconotext as a theoretical framework, the analysis combines semiotic examination of images and text with contextualization in early 20th-century rural Sweden. Social class is understood through Pierre Bourdieu´s concept of habitus as a generative system of durable dispositions that shape socially recognizable practices, tastes, and forms of behavior, through which individuals both enact and perceive classed positions within the social space. Maria Nikolajeva’s methodological insights on picture books provided a structured approach for examining environment, character, narrative perspective, time, movement, causality, and mimesis, without imposing a rigid sequence. The results show that idyll, in this book, is constructed through the interaction of setting and characters’ emotional states, and that it is dependent on stability, care, and adult responsibility, which in turn correlate with higher social class. Social differences are communicated through material culture, behavior, and interactions, with safety and trust contingent on the adults surrounding the child. In this context and in this book, idyll can be understood as dependent on security and stability, which in turn is closely linked to and considered contingent upon higher social class
Regarding the Use of AI in Social Studies Education : A Qualitative Interview Study
No full textDenna studie undersöker hur gymnasielärare i samhällskunskap uppfattar och använder artificiellintelligens (AI) i sin undervisning samt hur tekniken påverkar relationen mellan ämnesinnehåll,pedagogik och teknik. Studien tar sin utgångspunkt i TPACK-modellen, med särskilt fokus på lärarnasteknologiska pedagogiska kunskap (TPK) och teknologiska ämneskunskap (TCK). Empirin bygger påsex kvalitativa, semistrukturerade intervjuer med verksamma samhällskunskapslärare på gymnasiet.Resultatet visar att AI används på olika sätt och i varierande grad. För flera lärare fungerar AI främstsom ett stöd i planering, exempelvis vid skapande av undervisningsmaterial, medan andra i litenutsträckning använder tekniken på grund av begränsningar i skolans tillåtna verktyg. AI användsdärför oftare som en källa till inspiration än som ett verktyg som förändrar undervisningen i grunden.Studien visar också att AI påverkar ämnesinnehållet i samhällskunskap, särskilt inom områden somkällkritik och informationshantering. Lärarna beskriver att AI ibland ger förenklade eller felaktigasvar, vilket ökar behovet av arbete med kritiskt tänkande och förståelse för hur generativ AI fungerar.Lärarna beskriver elevers användning av AI som både en möjlighet och en utmaning. Enligt lärarnakan skillnader mellan elever förstärkas, beroende på hur väl eleverna förstår och använder tekniken.Dessa uppfattningar har lett till att flera lärare förändrat sina uppgifter och bedömningsformer, medökat fokus på muntliga examinationer, seminarier och bedömningssituationer som sker i klassrummet.Sammantaget visar studien att AI upplevs som en teknik som både skapar möjligheter och väcker nyadidaktiska frågor. Med TPACK som analytiskt ramverk synliggörs hur AI påverkar planering,undervisning, ämnesinnehåll och bedömning, samtidigt som tekniken ställer krav på lärarensprofessionella omdöme och etiska ansvar. Resultatet pekar på ett behov av fortsatt forskning ochtydligare stödstrukturer för att AI ska kunna användas på ett sätt som stärker både undervisningenskvalitet och skolans demokratiska uppdrag