BIBSYS: Open Journals Systems
Not a member yet
393 research outputs found
Sort by
Automatic translation from FBD-PLC-programs to NuSMV for model checking safety-critical control systems
Programmable logic
controllers (PLCs) are digital control systems, commonly used in industrial automation and
safety-critical applications. Control systems used in
safety-critical areas must undergo an extensive and
thorough certification and verification process. In
safety-critical applications, the PLC programming
standard IEC 61131-3 is widely accepted in
industry. PLC programmers who develop control
systems for safety-critical systems are often required
to verify the logic of PLCs by using formal methods
such as model checking. Translating manually from a
PLC program to the input language of a model checker
takes times and is often error-prone.
We develop a compiler to automatically translate PLC programs in the function block diagram (FBD) language, one of five industry standard PLC programming notations, to the input language of the model checker NuSMV. We have evaluated correctness, robustness, and performance of the PLC-NuSMV compiler using a case study. Evaluation results show that the compiler can translate the PLC programs correctly. The compiler can also identify several input errors and can scale to relative large PLC programs
Is the lecture dead?
Today, technology such as the Internet and automated digital video cameras offers the possibility of using new methods for teaching. We see that massive online open courses (MOOCs) and the idea of a “flipped classroom” are promising to replace the traditional lecture. The idea is to use technology to reduce costs and improve learning. However, in the quest for new teaching methods we have perhaps not understood why the lecture format is so widely used. In this paper, we shall try to explore the advantages of the traditional lecture
ADIOS LDA: When Grammar Induction Meets Topic Modeling
We explore the interplay between grammar induction and topic modeling approaches to unsupervised text processing. These two methods complement each other since one allows for the identification of local structures centered around certain key terms, while the other generates a document wide context of expressed topics. This approach allows us to access and identify semantic structures that would be otherwise hardly discovered by using only one of the two aforementioned methods. Using our approach, we are able to provide a deeper understanding of the topic structure by examining inferred information structures characteristic of given topics as well as capture differences in word usage that would be hard by using standard disambiguation methods. We perform our exploration on an extensive corpus of blog posts centered around the surveillance discussion, where we focus on the debate around the Snowden affair. We show how our approach can be used for (semi-) automated content classification and the extraction of semantic features from large textual corpora
Seadrive - Remote file synchronization for offshore fleets (short paper)
Seadrive is a novel file synchronizations framework and file hosting service for offshore naval fleets. Offshore vessels outside the range of cellular networks employ volatile satellite-based networks unsuited for file synchronization. Seadrive aspires to provide naval fleets with usable file hosting services over low bandwidth, high latency network links with high loss rate and large variability in the overall network stability. We provide the rationale for Seadrive; present the architecture, and design of the framework. We present an initial evaluation of the correctness and performance, preliminary results shows that Seadrive outperforms existing system in the test environment
Errors as Data Values
A “thrown” exception is a non-local side effect that complicates static reasoning about code. Particularly in functional languages it is fairly common to instead propagate errors as ordinary values. The propagation is sometimes done in monadic style, and some languages include syntactic conveniences for writing expressions in that style. We discuss a guarded- algebra-inspired approach for integrating similar, implicit error propagation into a language with “normal” function application syntax. The presented failure management approach accommodates language designs with all- referentially-transparent expressions. It furthermore supports automatically checking data invariants and function pre- and post-conditions, recording a trace of any due-to-an-error unevaluateable or failed expressions, and in some cases retaining “bad” values for potential use in recovering from an error
A Measurement-Driven Approach to Understand Urban Greenhouse Gas Emissions in Nordic Cities
Cities are main drivers for climate change mitigation and emission reduction today. However, in many cases they lack reliable baselines of emissions to validate current developments over time, assess the impact of their projects, and prioritize investments and actions. They also need better data on a small geospatial and temporal scale to really understand local emissions. This paper describes the rationale and the design of the Carbon Track and Trace project (CTT) that aims to develop an automated system for greenhouse gas (GHG) emissions monitoring through a low-cost city-level sensor network. The system is based on a flexible architecture incorporating open source sensor platforms, an Internet-of-Things wireless backbone, and extensive data analytics. We describe concept, architecture, and deployment as well as initial results
Bruk av norsk talegjenkjenning for virtuelle besøk innen helsesektoren
Artikkelen tar for seg hvordan virtuelle besøk kan nyttes innen telemedisin, og hvordan en kan ha nytte av talegjenkjenning pa? dette omra?det. En web-applikasjon er utviklet for a? utføre videosamtaler der talegjenkjenning brukes til diktering av det som blir sagt. Web-applikasjonen er ment for samtaler mellom pasient og helsepersonell. For a? oppna? talegjenkjenning ble det utviklet ba?de akustiske modeller (skjulte Markov modeller–HMM) og spra?kmodeller (Trigram) ved hjelp av a?pne tilgjengelige ressurser. Web-applikasjonen ble testet ut av sykepleierstudenter ved Høgskolen i Bergen. Sykepleierstudentene deltok ogsa? i en spørreundersøkelse for a? kartlegge studentenes meninger om nytten av virtuelle besøk.Ressursene som ble brukt for a? utvikle en talegjenkjenner pa? norsk var imidlertid ikke tilstrekkelige for bruk av diktering i web-applikasjonen (Word Error Rate (WER>100 %)). Studentene var positive til bruk av diktering, men hadde ulike meninger om nytten av virtuelle besøk
Gender Differences in Computer Science Education: Lessons Learnt from an Empirical Study at NTNU
Research in Computer Science (CS) education has focused on gender differences and there is high interest in increasing female participation in CS. The present study reviews important factors that influence CS students’ decision to complete their studies in CS (retention), and empirically examines how these factors differ for males and females. To this end, we identify cognitive and non-cognitive gains, cognitive and affective engagement, motivation to study, and three different barriers as critical factors in CS students’. We test these factors on 236 Norwegian CS students in order to find any potential gender differences. The findings indicate a gender difference for cognitive gains, affective engagement, motivation to study, and satisfaction with learning effectiveness. On the other hand, no difference was found for non-cognitive gains, cognitive engagement, personal values/teaching quality, and students’ intention to dropout from their studies. The study concludes with implications that will guide both research and practice towards a better understanding of both male and female CS students
Programmering som et pedagogisk verktøy i matematikkundervisningen
Programmering i skolen er et høyst aktuelt tema, og interessen for forskning på dette temaet er økene. I dette prosjektet har vi utviklet og fulgt et undervisningsopplegg ved to skoler i Østfold, hvor programmering har blitt testet i matematikkundervisningen. Elevene på skolene tok i bruk programmeringsverktøy for å løse matematiske oppgaver og problemer. Dette prosjektet fungerer som en forstudie for å finne ut hvordan programmering ble tatt i mot av elever og lærere på de to skolene. Det konkluderes med at programmering i matematikktimene kan være en annerledes måte å løse matematiske problemer, og gi elevene et annet perspektiv på matematikk. Derimot er det faktorer som bør vurderes hvis programmering skal innføres som et pedagogisk verktøy skolen
Training Googles SyntaxNet to understand Norwegian Bokmål and Nynorsk (short paper)
We use Google’s open source neural network framework, SyntaxNet, to train a fully automatic part-of-speech language tagger for Norwegian Bokmål and Nynorsk. Using SyntaxNet, we are able to get comparable results with other tagger when tagging Bokmål and Nynorsk with part-of-speech. Both taggers are available and released as open source