BIBSYS: Open Journals Systems
Not a member yet
393 research outputs found
Sort by
Programmeringsoppgaver og auto-scoring: Inspera vs. andre løsninger
Full text linkRetting av store antall eksamensbesvarelser i programmeringsfag innen stramme sensurfrister er krevende. Selv om oppgavetypen «Programmering» i Inspera gir klare fordeler framfor håndskrevet kode, er den manuelt rettet. I 2021 lanserte Inspera en ny oppgavetype, «Kompilering», hvor poeng kunne gis automatisk basert på testsuiter. Dessverre var oppgavetypen beheftet med en del problemer, og den er nå trukket tilbake for videreutvikling, dermed ikke lenger tilgjengelig for bruk i eksamen. Denne artikkelen beskriver våre erfaringer med oppgavetypen «Kompilering» og diskuterer hvilke behov for forbedringer som ville være ønskelig for at bruk av denne skulle kunne bli vellykket - særlig for eksamen, men også potensielt for bruk under formativ vurdering underveis i semesteret. Det gjøres en sammenligning med andre systemers implementasjon av lignende oppgavetyper, f.eks. Moodle CodeRunner, basert på praktisk utprøving og litteraturstudium av erfaringer med bruk i andre land. Artikkelen diskuterer også i hvilken grad karaktersetting kan basere seg fullt ut på automatisk poenggivning av kode, eller hvorvidt man fortsatt vil trenge manuell gjennomgang, for eksempel for å kunne gi noe poeng til studenter som har kode som ikke fungerer, men hvor løsningen likevel inneholder mye riktig tankegang
Analysing the feasibility of using Objectosphere for Face Presentation Attack Detection
No full textSee paper
Developing a Corequisite Writing Textbook: How Two Novices Handled the Complex Nature of Open Textbook Production
Full text linkDespite the growing body of scholarship focused on Open Educational Resources (OERs), studies tracking open textbook production are exceedingly rare. To gain insight into the complexities associated with open textbook development, this article’s authors used concurrent verbalization and pre- and post-concurrent verbalization interviews to document writing episodes while composing their first coursebook, an open-source text designed for corequisite course pairings of writing fundamentals and introduction to technical writing. Corequisite classes combine content-area instruction with explicit skill-building opportunities, and although commercial publishing houses do produce corequisite textbooks for traditional general education courses, the authors were impelled by the need to create a specialized coursebook for a STEM-focused writing corequisite. Qualitative content analysis of the data collected during ongoing coursebook development revealed how the novice textbook authors navigated complexities when sourcing material for the book, formatting content, and embedding hyperlinks into chapters. Open textbook development offers a variety of professional development opportunities, and the details provided herein could be instructive to others embarking on open coursebook projects and inspire further inquiry into ongoing open textbook production. This paper also exposes the dynamic interaction between textbook content as designed and operationalized, offering implications for the research field of materials development
Dybdelæring, tverrfaglighet og samskaping – en prosessuell studie av samarbeid mellom IT- og tjenestedesignstudenter
Full text linkI denne artikkelen presenterer vi funn og erfaringer fra et tverrfaglig samarbeid i bachelorstudier ved Universitetet i Sørøst-Norge, mellom første års IT-studenter ved IT-utdanningen og tredje års tjenestedesignstudenter ved økonomi- og ledelsesutdanningen. Våre mål med samarbeidet var å bidra til dybdelæring for studentene, samt gi dem erfaring med tverrfaglig samarbeid slik de vil erfare i arbeidslivet. Vi har observert samarbeidsprosessene underveis, og foretatt uformelle samtaler med IT- og tjenestedesignstudenter fra to årskull og til sammen fire ulike klasser. Erfaringene fra samhandling er blandede. Mange studenter rapporterte godt læringsutbytte og dybdelæring, mens andre var mer kritiske til hva de fikk ut av det tverrfaglige samarbeidet. Koronasituasjonen med stengte campus og omlegging til digital kommunikasjon og samarbeid bidro i stor grad til blandede resultater. Vi avrunder artikkelen med å presentere et sett læringspunkter for lignende samarbeidsprosjekter i høyere utdanning
Selecting a GC for Java Applications
Full text linkNowadays, there are several Garbage Collector (GC) solutions that can be used in an application. Such GCs behave differently regarding several performance metrics, in particular throughput, pause time, and memory usage. Thus, choosing the correct GC is far from trivial due to the impact that di?erent GCs have on several performance metrics. This problem is particularly evident in applications that process high volumes of data/transactions especially, potentially leading to missed Service Level Agreements (SLAs) or high cloud hosting costs. In this paper, we present: i) thorough evaluation of several of the most widely known and available GCs for Java in OpenJDK HotSpot using different applications, and ii) a method to easily pick the best one. Choosing the best GC is done while taking into account the kind of application that is being considered (CPU or I/O intensive) and the performance metrics that one may want to consider: throughput, pause time, or memory usage
BAT: A Benchmark suite for AutoTuners
Full text linkthe code by ?nding the best possible values for a given architecture. To our knowledge, there are currently no standardized benchmark suites for comparing and testing autotuners. Developers of autotuners thus make their own when presenting and comparing autotuners. We thus present BAT, a Benchmark suite for AutoTuners with HPC-based parameterized GPU programs. CUDA programs and kernels from "The Scalable Heterogeneous Computing (SHOC) Benchmark" are parameterized. BAT contains a varied selection of benchmarks of different complexity that can utilize multiple GPUs on one system, either by running the same program and computations on multiple nodes, or by splitting the work between nodes. BAT contains 9 di?erent HPC benchmarks that provide a large search space of autotuning parameters, and are modified to suite many di?erent autotuners. BAT also includes a CLI that facilitates autotuning with the benchmarks. Our benchmark suite is tested with four di?erent autotuners, OpenTuner, Kernel Tuner, CLTune and KTT. They di?er in setup and how they tune. The impact of the di?erent benchmark parameters on the running time across architectures is analyzed. Test systems used include a DGX-2, IBM Power System AC922 with Tesla V100-SXM2 32 GB GPUs, an RTX Titan, a GeForce GTX 980 and a server with 20 Tesla T4 GPUs
Introduksjonsdag for gruppeledere i informatikk: hvilke tema bør adresseres?
Full text linkPoster submissio
Desktop Video Conferencing Tools in HigherEducation: Understanding Lecturers’ Experience
Full text linkTechnological advancements facilitate new ways of conducting university lectures. Through Emergency Remote Teaching (ERT), Desktop Video Conferencing Tools (DVCT) have been adopted by higher education institutions to face the challenges of the COVID-19 pandemic. DVCT enables online, real-time education to be delivered. To understand educators’ experience with DVCT, an online survey was con-ducted among lecturers (N = 243) at the University of Bergen (UiB) between October 12-29th 2020. In this empirical paper, we report on the findings from the Likert scale and free text questions, analyzed using quantitative and qualitative approaches. Elaborate qualitative data (18,107 words) was gathered, providing perspectives of lecturers’ experiences. The main advantages of DVCT identified are flexibility and accessibility. Disadvantages concern poor student communication and interaction. Overall, the trend is mixed to low sentiments towards lecturing with DVCT, despite lecturers claiming to be confident in using new digital tools. Lecturers’ experience with DVCT may be positively affected by providing them with training, emphasizing pedagogical aspects, accepting the shortcomings of ERT, and recognizing that digitalization encompasses adapting to new technologies as well as changing practices, planning, and execution
Toward a decision support system for COVID-19 vaccine allocation inside countries
Full text linkThe distribution of COVID-19 vaccines has proved to be a challenging task for public health authorities in many countries. Among several decisions involved in the task, allocating limited available vaccines to administration points is indeed critical. However, the operation management literature lacks evidence-based mathematical models that could support effective, efficient, sustainable and equitable vaccine allocation decision. This paper develops the fundamentals of a decision support system for COVID-19 vaccine allocation inside countries. The proposed DSS intends to support public health authorities in real-time by illustrating possible vaccine alternatives. The system could also inform and support other actors in the COVID19 distribution for planning and collaboration. Two illustrative cases for the COVID-19 vaccine allocation have been investigated to highlight potential benefits of our methodology
Knowledge and Skills Needed to Craft Successful Cybersecurity Strategies
Full text linkDaily advancing technologies and next-generation networks are creating entirely dierent digital environments for people, organizations, and governments within the next several years. Because cybersecurity provision in such environments involves many actors and must overcome many evolving threats and challenges, strategies must be responsive and multi-pronged. Development and execution of suficiently savvy strategies to face the complex problems in this context necessitate identification of all the actors and operations that aect, directly or indirectly, on the cybersecurity of the digital ecosystems. In this study, we seek to provoke thinking about how actors and stakeholders could get better at crafting successful cybersecurity strategies, and identify and integrate specic types of skills required to formulation these strategies taking into account where decisions are actually made. This work provides an insight into cybersecurity education, calibrating and differentiating knowledge and skills to make the right demands on the right actors who have the authority and responsiveness to introduce change from multiple entry points. This enables practitioners to adopt more hands-on approaches that can be helpful to improve transparency, accountability and collaboration across levels of a socio-technical system