BIBSYS: Open Journals Systems
Not a member yet
393 research outputs found
Sort by
Improved Graph Indexing Algorithms for Label-Constrained Reachability Queries
Get PDFNowadays graph data have become absolutely ubiquitous in various applications starting from social/road networks to bio-medical data etc. Given such graph data, a reachability query asks if there exists a path from a source vertex to a target vertex in the graph. Due to its immense implications in both theory and applied domains, this query and many of its variants have been extensively studied in the literature. One such variant investigates the reachability between two vertices in an edge-labeled graph while constraining the label set simultaneously. This problem has recently been addressed by Valstar et al. [SIGMOD'17] who proposed an approach called the landmark indexing (LI) to support faster label-constrained reachability (LCR) queries. In this work, we introduce a simple, practical and space-e?cient solution for answering LCR queries even faster. The experimental evaluation shows signi?cant time and space e?ciency bene?ts of our proposed solution over the LI approach for this problem in both real-world and synthetic graphs
Hyperspectral imaging and machine learning for the prediction of SSC in kiwi fruits
Get PDFSolids Content (SSC) of the fruits in a non-destructive way. A database is created which includes the hyperspectral data acquired in the visible and near-infrared region (VNIR) and measurements done with a sugar meter.We have applied di?erent machine learning techniques to investigate the correlation between spectral information and the SSC. The models tested were support vector regression (SVR), k-nearest neighbor (KNN), partial least squares (PLS), and multiple linear regression (MLR) with di?erent variable selection techniques and dimensionality reduction. The best model at determining SSC was Uninformative Variable Elimination (UVE)-PLS, which had RMSE = 1.047 °Brix and R2 = 0.39 on the test set
Polygonal blending splines in application to image processing
Get PDFThe paper proposes a novel method of image representation. The basic idea of the method is to transform color images to continuous
parametric surfaces. The proposed technique is based on a class of special basis functions, defined on the polygon grid. Besides a exible and symmetric construction, these basis functions are strictly local and Cd-smooth on the entire domain. Having a number of unique features, the proposed representation can be used in various image processing tasks. The main purpose of this paper is to demonstrate the process of the image transformation and discuss possible applications of the presented technique
Ph.D. Academic Life at the Time of the Coronavirus
Get PDFPh.D. students are an essential part of the academic world. They need to obtain new knowledge and skills while contributing to the advancement of their re-search field. This paper examines the way the academic life of Ph.D. students is affected by the pandemic COVID-19. Data for this study was collected through an online survey sent to the Ph.D. students of a department in a Scandinavian university. Data was then analyzed through a mixed (quantitative and qualitative) method. The questionnaire was distributed after one year of studying and working from the home office and far from an educational setup like physical classes, meetings, seminars, and conferences. The results show changes in how they meet their supervisors and other Ph.D. students and the type of tools they use for communications. The results also show that Ph.D. students face somechallenges related to the home office, project-related activities like data collection and experiment, and contact with others like socializing, communication, and collaboration. To reduce the impact of the Coronavirus, Ph.D. studentsalso suggested some solutions like having more online discussions with their supervisors and peer Ph.D. students, having a work-life balance, getting guidance and training from professionals, and being understood by the supervisors
Picture perfect: ELT textbook images and communicative competence development
Get PDFImages are an important resource in English language teaching (ELT) textbooks. Still, knowledge of how textbook images may support the development of the learner’s communicative competence is still scarce. This paper provides examples of how textbook images can support such a development. The theoretical framework consists of models of communicative competence, the notions of input and output, principles for communicative approaches, as well as key tools for image analysis. As examples relating to the phenomenon under study, Nicaraguan textbooks for the teaching of English in secondary school have been used
Referansegrupper som verktøy for kontinuerlig kvalitetsforbedring
Get PDFNTNU sitt kvalitetssystem stiller krav til at emneansvarlig skal sørge for kontinuerlig kvalitetsutvikling i emner de har ansvar for. Emneansvarlig skal normalt gjennomføre emneevaluering ved bruk av referansegruppemetodikk som grunnlag, for å dokumentere og rapportere kvalitetsarbeid og utviklingstiltak i sine emner. Vi ser fra litteraturen at det er bæde styrker og svakheter knyttet til bruken av referansegruppemetodikk. For å kartlegge dagens praksis og erfaringer med metodikken ved vårt fakultet på NTNU, har vi utarbeidet og gjennomført en spørre-undersøkelse rettet mot emneansvarlige og studenter ved flere av våre studieprogrammer. Med utgangspunkt i svarene fra spørreundersøkelsene og ulike artikler erfarer vi at emneevaluering og bruken av referansegrupper gjennomføres svært forskjellig og med ulikt opplevd utbytte. Resultater som vi har kommet fram til i denne artikkelen viser at bruk av referansegrupper som metode i kombinasjon med litt større fokusgrupper for evaluering av emner, kan gi verdifulle bidrag inn i, og sikre kontinuitet i kvalitetsarbeidet. Studenter og emneansvarlig opplever over tid kvalitetsutvikling og konkrete endringer basert på arbeidet som er gjort i referansegruppene. Videre erfarer vi at både studenter og emneansvarlige er positivt innstilt til emneevaluering og referansegruppemetodikk - under gitte forutsetninger - der den enkeltfaktoren som har størst betydning for åade gjennomføring og resultat er emneansvarliges engasjement til evalueringsarbeidet
Git in an Educational Context
Get PDFThe version control system Git is commonly used in computer science education. Best practice use of Git is a relevant professional skill that students need to learn, but Git also is a natural choice for efficient collaboration on programming assignments. The learning context is however different from the professional context. In this paper we explore the implications Git has in an educational context using a questionnaire and result implies that Git influences some students’ experience with learning and group work. However, it also verifies previous findings that using Git is beneficial for student collaboration. Data from the use of Git can potentially inform students and educators about aspects of group work. A prototype mirroring tool using GitLab data has been developed to investigate the data’s potential and semi-structured interviews with students and teachers were conducted to evaluate and explore the tool. Results suggest that GitLab data is well-suited to provide students new insight into their work and educators with an efficient method for monitoring project work and student groups
The News Angler Project: Exploring the Next Generation of Journalistic Knowledge Platforms
Get PDFThe News Angler project aims to support journalists in finding new and unexpected connections and angles in the news. The project therefore explores how recent artificial intelligence (AI) techniques — such as knowledge graphs, natural-language processing (NLP) and machine learning (ML) — can support high-quality journalism that exploits big and open data sources. A central contribution is News Hunter, a series of prototype journalistic knowledge platforms (JKPs)
The Role of Citizens’ Familiarity, Privacy Concerns, and Trust on Adoption of Smart Services
Get PDFSmart city solutions and applications are considered as a strategic means to cope with multiple global and local challenges such as pollution, energy expenditure and digitalization to name a few. Although these solutions are driven by advanced information technologies such as IoT and Big data, their success is dependent on user engagement and trust. We seek to examine how citizens' awareness and perception of smart cities affect their adoption of smart services.To answer this, we conducted a study in Norway and employed a questionnairereceiving 103 responses. Furthermore, we conducted 12 semi-structured interviewsto obtain further insights. The results show how citizens value the benefitof smart services and how their adoption is influences by engagement and trusttowards them.
Morphing Attack Detection using Laplace operator based features
Get PDFThe vulnerability of facial recognition systems through morphing attacks is a known problem. Since the first publication about this vulnerability of facial recognition systems, a variety of morphing attack detection methods have been presented, promising an automated detection of such fraudulent attacks. In this work, a new approach is presented attempting to distinguish bona fide from morphed images based on information about the edges in the image extracted by the Laplace operator. It can be demonstrated that the features employed contain information that can contribute to the detection of morphed face images