Computer Science Journal (AGH University of Science and Technology, Krakow)
Not a member yet
476 research outputs found
Sort by
Challenges and Opportunities for the Future of iCampuses
Get PDFMeeting the educational needs of students currently requires moving towardcollaborative electronic and mobile learning systems that parallel the vision ofWeb 2.0. However, factors such as data freedom, brokerage, interconnectivityand the Internet of Things add to a vision for Web 3.0 that will require con-sideration in the development of future campus-based, distance and vocationalstudy. So, education can, in future, be expected to require deeper technologicalconnections between students and learning environments, based on significantuse of sensors, mobile devices, cloud computing and rich-media visualization.Therefore, we discuss challenges associated with such a futuristic campus con-text, including how learning materials and environments may be enriched byit. As an additional novel element the potential for much of that enrichmentto be realized through development by students, within the curriculum, is alsoconsidered. We will conclude that much of the technology required to embracethe vision of Web 3.0 in education already exists, but that further research inkey areas is required for the concept to achieve its full potential
IMPROVEMENTS TO GLOWWORM SWARM OPTIMIZATION ALGORITHM
Get PDFGlowworm Swarm Optimization algorithm is applied for the simultaneous capture of multipleoptima of multimodal functions. The algorithm uses an ensemble of agents, which scan thesearch space and exchange information concerning a fitness of their current position. Thefitness is represented by a level of a luminescent quantity called luciferin. An agent movesin direction of randomly chosen neighbour, which broadcasts higher value of the luciferin.Unfortunately, in the absence of neighbours, the agent does not move at all. This is anunwelcome feature, because it diminishes the performance of the algorithm. Additionally,in the case of parallel processing, this feature can lead to unbalanced loads. This paperpresents simple modifications of the original algorithm, which improve performance of thealgorithm by limiting situations, in which the agent cannot move. The paper provides resultsof comparison of an original and modified algorithms calculated for several multimodal testfunctions
NORMAL FUNCTIONS AS A NEW WAY OF DEFINING COMPUTABLE FUNCTIONS
Get PDFReport sets new method of defining computable functions. This is formalization of traditional function descriptions, so it allows to define functions in very intuitive way. Discovery of Ackermann function proved that not all functions that can be easily computed can be so easily described with Hilbert’s system of recursive functions. Normal functions lack this disadvantage
Staged event-driven architecture as a micro-architecture of distributed and pluginable crawling platform
Get PDFThere are many crawling systems available on the market but they are rather close systems dedicated for performing particular kind and class of tasks with predefined set of scope, strategy etc. In real life however there are meaningful groups of users (e.g. marketing, criminal or governmental analysts) requiring not just a yet another crawling system dedicated for performing predefined tasks. They need rather easy-to-use, user friendly all-in-one studio for not only executing and running internet robots and crawlers, but also for (graphical) (re)defining and (re)composing crawlers according to dynamically changing requirements and use-cases. To realize the above-mentioned idea, Cassiopeia framework has been designed and developed. One has to remember, however, that enormous size and unimaginable structural complexity of WWW network are the reasons that, from a technical and architectural point of view, developing effective internet robots – and the more so developing a framework supporting graphical robots’ composition – becomes a really challenging task. The crucial aspect in the context of crawling efficiency and scalability is concurrency model applied. There are two the most typical concurrency management models i.e. classical concurrency based on the pool of threads and processes and event-driven concurrency. None of them are ideal approaches. That is why, research on alternative models is still conducted to propose efficient and convenient architecture for concurrent and distributed applications. One of promising models is staged event-driven architecture mixing to some extent both of above mentioned classical approaches and providing some additional benefits such as splitting application into separate stages connected by events queues – what is interesting taking requirements about crawler (re)composition into account. The goal of this paper is to present the idea and the PoC implementation of Cassiopeia framework, with the special attention paid to its crucial architectural element i.e. design, implementation and applying of staged event-driven architecture being a micro-architecture of Cassiopeia’s agents i.e. its key computational and processing unit
HIGH ENERGY PHYSICS EXPERIMENTS IN GRID COMPUTING NETWORKS
Get PDFThe demand for computing resources used for detector simulations and data analysis in HighEnergy Physics (HEP) experiments is constantly increasing due to the development of studiesof rare physics processes in particle interactions. The latest generation of experiments at thenewly built LHC accelerator at CERN in Geneva is planning to use computing networks fortheir data processing needs. A Worldwide LHC Computing Grid (WLCG) organization hasbeen created to develop a Grid with properties matching the needs of these experiments. Inthis paper we present the use of Grid computing by HEP experiments and describe activitiesat the participating computing centers with the case of Academic Computing Center, ACKCyfronet AGH, Kraków, Poland
Two-dimensional HP-adaptive Algorithm for Continuous Approximations of Material Data Using Space Projection
Get PDFIn this paper we utilize the concept of the L2 and H1 projections used toadaptively generate a continuous approximation of an input material data inthe finite element (FE) base. This approximation, along with a correspondingFE mesh, can be used as material data for FE solvers. We begin with a brieftheoretical background, followed by description of the hp-adaptive algorithmadopted here to improve gradually quality of the projections. We investigatealso a few distinct sample problems, apply the aforementioned algorithms andconclude with numerical results evaluation
GRAPH IMAGE LANGUAGES IN SEMANTIC DESCRIPTION OF SPATIAL CORONARY ARTERIES STRUCTURE
Get PDFIn this paper has been proposed developing the new syntactic – semantic meaning descriptionof spatial coronary arteries structure. Thanks such description will be possible to makeessentially steered semantic interpretation of section coronary arteries morphology, what willallow us fast identification and automatisation of lumen stricture detection. In this aim hasbeen used graph image languages based on the expansive graph grammars of edNLC type,enabling creation the universal and informative meaning description of spatial coronaryarteries structure. Application of such semantic description in the integrated modules ofintelligent systems medical diagonosis, supporting the early detection stricture which defectthe flow of oxidizing blood to given area of cardiac muscle
AUTOMATIC CONTEXTUAL TEXT CORRECTION USING THE LINGUISTIC HABITS GRAPH LHG
Get PDFAutomatic text correction is an essential problem of today text processors and editors. Thispaper introduces a novel algorithm for automation of contextual text correction using a LinguisticHabit Graph (LHG) also introduced in this paper. A specialist internet crawler hasbeen constructed for searching through web sites in order to build a Linguistic Habit Graphafter text corpuses gathered in polish web sites. The achieved correction results on a basis ofthis algorithm using this LHG were compared with commercial programs which also enableto make text correction: Microsoft Word 2007, Open Office Writer 3.0 and search engineGoogle. The achieved results of text correction were much better than correction made bythese commercial tools
Modular FEM framework "ModFem" for generic scientific parallel simulations
Get PDFWe present the design and its\u27 implementation for a flexible and robust parallel modular finite element (FEM) framework, called ModFem. The design is based on reusable modules which use narrow and well-defined interfaces to cooperate. At the top of the architecture there are problem dependent modules. Problem dependent modules can be additionally grouped together by "super-modules". This structure allows for applying the sequential codes to parallel environments and also support solving multi-physics and multi-scale problems
NEURAL NETWORKS FOR ELECTROHYDRODYNAMIC EFFECT MODELLING
Get PDFThis paper presents currently achieved results concerning methods of electrohydrodynamiceffect used in geophysics simulated with feedforward networks trained with backpropagation algorithm, radial basis function networks and generalized regression networks