Computer Science Journal (AGH University of Science and Technology, Krakow)
Not a member yet
476 research outputs found
Sort by
Policy-based SLA storage management model for distributed data storage services
Get PDFThere is high demand for storage related services supporting scientists in their research activities. Those services are expected to provide not only capacity but also features allowing for more flexible and cost efficient usage. Such features include easy multiplatform data access, long term data retention, support for performance and cost differentiating of SLA restricted data access. The paper presents a policy-based SLA storage management model for distributed data storage services. The model allows for automated management of distributed data aimed at QoS provisioning with no strict resource reservation. The problem of providing users with the required QoS requirements is complex, and therefore the model implements heuristic approach for solving it. The corresponding system architecture, metrics and methods for SLA focused storage management are developed and tested in a real, nationwide environment
Towards Trasparent Data Access with Context Awareness
Get PDFApplying the principles of open research data is an important factor accelerating the production, analysis of scientific results and worldwide collaboration. However, still very little data is being shared. The aim of this article is analysis of existing data access solutions in order to identify reasons for such situation. After analysis of existing solutions and data access stakeholders needs, the authors propose own vision of data access model evolution
Using Erlang in Research and Education in a Technical University
Get PDFThis paper addresses the problem of using functional programming (FP) languages for research and education purposes. In order to identify problems associated with usage of FP languages, such as Erlang, an experiment consisting of two surveys was performed. The rst survey was anonymous, and aimed at establishing whether the participants prefer object-oriented or functional coding. The second one was a survey after students have nished an Erlang course. The results of these two surveys demonstrate that functional programming is underrated without apparent reasons. Possible steps to address this problem are suggested
A new approach to storing dynamic data in relational databases using JSON
Get PDFJavaScript Object Notation was originally designed to transfer data; however, it soon found another use as a way to persist data in NoSQL databases. Recently, the most popular relational databases introduced JSON as native column type, which makes it easier to store and query dynamic database schema. In this paper, we review the currently popular techniques of storing data with a dynamic model with a large number of relationships between entities in relational databases. We focus on creating a simple dynamic schema with JSON in the most popular relational databases and we compare it with well-known EAV/CR data model and the document database. The results of precisely selected tests in the field of Criminal Data suggest that the use of JSON in dynamic database schema greatly simplifies queries and reduces their execution time compared to widely used approaches
Emergence of population structure in socio-cognitively inspired ant colony optimization
Get PDFA metaheuristic proposed by us recently, Ant Colony Optimization (ACO) hybridized with socio-cognitive inspirations, turned out to generate interesting results compared to classic ACO. Even though it does not always find better solutions to the considered problems, it usually finds sub-optimal solutions usually. Moreover, instead of a trial-and-error approach to configure the parameters of the ant species in the population, in our approach, the actual structure of the population emerges from predefined species-to-species ant migration strategies. Experimental results of our approach are compared against classic ACO and selected socio-cognitive versions of this algorithm
EFFICIENT APPROACH FOR VIEW SELECTION FOR DATA WAREHOUSE USING TREE MINING AND EVOLUTIONARY COMPUTATION
Get PDFSelection of a proper set of views to materialize plays an important role indatabase performance. There are many methods of view selection which uses different techniques and frameworks to select an efficient set of views for materialization. In this paper, we present a new efficient, scalable method for view selection under the given storage constraints using a tree mining approach and evolutionary optimization. Tree mining algorithm is designed to determine the exact frequency of (sub)queries in the historical SQL dataset. Query Cost model achieves the objective of maximizing the performance benefits from the final view set which is derived from the frequent view set given by tree mining algorithm. Performance benefit of a query is defined as a function of queryfrequency, query creation cost, and query maintenance cost. The experimental results shows that the proposed method is successful in recommending a solution which is fairly close to optimal solution
A multi-agent blackboard architecture for supporting legal decision-making
Get PDFOur research objective is to design a system to support legal decision-making using the multi-agent blackboard architecture. Agents represent experts that may apply various knowledge processing algorithms and knowledge sources. Experts cooperate with each other using blackboard to store facts about current case. Knowledge is represented as a set of rules. Inference process is based on bottom-up control (forward chaining). The goal of our system is to find rationales for arguments supporting different decisions for a given case using precedents and statutory knowledge. Our system also uses top-down knowledge from statutes and precedents to interactively query the user for additional facts, when such facts could affect the judgment. The rationales for various judgments are presented to the user, who may choose the most appropriate one. We present two example scenarios in Polish traffic law to illustrate the features of our system. Based on these results, we argue that the blackboard architecture provides an effecive approach to model situations where a multitude of possibly conflicting factors must be taken into account in decision making. We briefly discuss two such scenarios: incorporating moral and ethical factors in decision making by autonomous systems (e.g. self-driven cars), and integrating eudaimonic (well-being) factors in modeling mobility patterns in a smart city
Model for Dynamic and Hierarchical Data Repository in Relational Database
Get PDFThe aim of this research is to build an open schema model for digital sources repository in the relational database. It required us to develop a few advanced techniques. One of them was to keep and maintain a hierarchical data structure pushed into the repository. The second was to create constraints on any hierarchical level that allow enforcing data integrity and consistency. The created solution is mainly based on a JSON as a native column type, which was designed for holding open schema documents. In this paper, we present the model for any repository that uses hierarchical dynamic data. Additionally, we include a structure for normalizing input and description for data to keep all the model assumptions. We compared our solution with well-known open schema model -- Entity-Attribute-Value -- in the scope of saving data and querying about relationship and content from structure. Results have shown that we achieved improvement in both performance and disk space usage, although we extended our model with a few new features that the previous model does not include. The techniques developed in this research can be applied in every domain where hierarchical dynamic data is required, as demonstrated by the digital book repository that we have presented
A parallel algorithm of ICSYM for complex symmetric linear systems in quantum chemistry
Get PDFComputational effort is a common issue for solving large-scale complex symmetric linear systems, particularly in quantum chemistry applications. In order to alleviate this problem, we propose a parallel algorithm of improved conjugate gradient-type iterative (CSYM). Using three-term recurrence relation and orthogonal properties of residual vectors to replace the tridiagonalization process of classical CSYM, which allows to decrease the degree of the reduce-operator from two to one communication at each iteration and to reduce the amount of vector updates and vector multiplications. Several numerical examples are implemented to show that high performance of proposed improved version is obtained both in convergent rate and in parallel efficiency
Machine learning models for predicting patients survival after liver transplantation
Get PDFIn our work we have built models predicting whether a patient will lose an organ after liver transplantation within a specified time horizon. We have used the observations of bilirubin and creatinine in the whole first year after the transplantation to derive predictors capturing not only their static value but also variability. Our models indeed have predictive power which proves the value of incorporating variability of biochemical measurements and it is the first contribution of our paper.The second one is the selection of the best model for the defined problem. We have identified that full-complexity models, such as random forests and gradient boosting, despite having the best predictive power, lack sufficient interpretability which is important in medicine. We have found that generalized additive models (GAM) provide desired interpretability and their predictive power is closer to the predictions of full-complexity models than to the predictions of simple linear models