21 research outputs found
An Event-Driven Architecture for Genomics-Based Diagnostic Data Processing
No full textGenomics-based diagnostic data (GBDD) are becoming increasingly important for laboratory diagnostics. Due to the large quantity of data and their heterogeneity, GBDD poses a big data challenge. Current analysis tools for GBDD are primarily designed for research and do not meet the requirements of laboratory diagnostics for automation, reliability, transparency, reproducibility, robustness, and accessibility. This makes it difficult for laboratories to use these tools in tests that need to be validated according to regulatory frameworks and to execute tests in a time- and cost-efficient manner. In order to better address these requirements, we propose an event-driven workflow-based architecture as the basis for a processing platform that is highly scalable using container technologies and microservices. A prototype implementation of this approach, called GenomicInsights, has been developed and evaluated to demonstrate its feasibility and suitability for laboratory diagnostics
Nutzerbefähigung in der Analyse von Big Data durch Anwendung von künstlicher Intelligenz und maschinellem Lernen
Get PDFGlobal data volume is growing exponentially, driven by manifold reasons: Smartphones and mobile connectivity, IoT, social media, and many others. Data availability drives progress in many application domains. Examples comprise pharmaceutical research, traceability of contagious diseases, research related to climate change, and data-driven insurance models. The rise of AI is closely connected to the availability of large data quantities. Although AI has been researched since the 1950s, it has recently gained significant momentum since it improves with more data available. AI and ML already affect everyday interactions with digital devices (e.g. face recognition) and will further penetrate all industrial sectors. Generative AI and Large Language Models already demonstrate how AI utilizes Big Data to solve complex tasks while lowering entry barriers and making it available to the general public. The growing scientific and economic relevance of Big Data, AI, and ML drives research, e.g. the European Commission-funded projects SMART VORTEX and EDISON. SMART VORTEX focused on defining Big Data-related infrastructure and creating a reference implementation. EDISON followed up on this by investigating the corresponding skill gap on the market and creating a competence framework (e.g. skill profiles and education plans). However, it did not investigate the concept of user-empowering Information Systems that lower entry barriers and reduce induction needs. Bornschlegl continued with the research on these challenges with his dissertation. He introduced IVIS4BigData, a reference model for visual Big Data analysis that enables standardized user stereotypes to participate in analyzing Big Data. None of these projects thoroughly considered AI and ML and their implications. However, AI and ML bear the potential to provide methods for analyzing Big Data and improving Big Data infrastructure and tooling. This requires defining AI and ML representation alongside the Big Data analysis pipeline and overcoming an intensified skill challenge, adding capability requirements for AI and ML. Architecture blueprints for implementing this integration do not exist. Thus, it remains a challenge to investigate AI and ML applications for Big Data-related infrastructure and tooling (SMART VORTEX), their impact on user capabilities (EDISON), and their integration into user-empowering Big Data analysis IS (IVIS4BigData). To close this gap, this dissertation developed AI2VIS4BigData, a reference model for end user-empowering AI and ML integration into the Big Data analysis pipeline that defines AI and ML representation, categorizes user stereotypes, and provides a service-oriented reference architecture. It was implemented for seven application domains (e.g. meteorology) and thoroughly evaluated, e.g. in workshops and studies.Die weltweite Datenmenge wächst aufgrund vielfältiger Ursachen wie Smartphones und Mobilfunktechnologie, dem Internet der Dinge und sozialen Netzwerken exponentiell. Die Verfügbarkeit von Daten treibt den Fortschritt in verschiedenen Anwendungsdomänen voran. Beispiele umfassen die Medikamentenforschung, die Überwachung ansteckender Krankheiten, Forschung mit Klimawandel-Bezug sowie datengetriebene Versicherungsmodelle. Die steigende Relevanz von KI ist eng mit der Datenverfügbarkeit verbunden. Auch wenn KI bereits seit den 1950er Jahren erforscht wird, so stieg in den letzten Jahren die Bedeutung von KI mit steigender Datenmenge auf Grund mit einer mit den Daten wachsenden Leistungsfähigkeit. KI und maschinelles Lernen beeinflussen schon heute das digitale Leben von jedem (bspw. durch Gesichtserkennung) und werden zukünftig alle industriellen Bereiche durchdringen. Generative KI und große Sprachmodelle (sog. Large Language Models) demonstrieren bereits, wie KI komplexe Aufgaben lösen und gleichzeitig die Hemmschwelle so senken kann, dass die Allgemeinheit daran partizipiert. Die wachsende wissenschaftliche und wirtschaftliche Relevanz von Big Data, KI und ML beschleunigt die Forschung bspw. durch die von der Europäischen Kommission geförderten Forschungsprojekte SMART VORTEX und EDISON. Der Schwerpunkt von SMART VORTEX lag auf der Definition von Big Data Infrastruktur und Referenzimplementierungen. EDISON hat mit einer Untersuchung der Bedarfslücke der notwendigen Nutzerfähigkeiten daran angeknüpft und ein Kompetenz-Rahmenwerk (z.B. Fähigkeitsprofile, Bildungspläne) geschaffen. Es hat jedoch keine nutzer-befähigenden Informationssysteme untersucht, die Eintrittshürden absenken und Einarbeitungsbedarfe reduzieren können. Bornschlegl setzte mit seiner Dissertation die Forschung an diesen Herausforderungen fort. Er hat mit IVIS4BigData ein Referenzmodell für visuelle Analyse von Big Data eingeführt, das standardisierten Nutzerstereotypen die Partizipation am Big Data-Analysevorgang ermöglicht. Keines dieser Projekte hat den Einsatz von KI und ML in hohem Maße berücksichtigt. Dabei haben KI und ML das Potenzial, als Big Data-Analysemethode eingesetzt werden zu können oder Big Data-Infrastruktur und -Tooling zu verbessern. Dazu müssen allerdings die Repräsentation von KI und ML entlang der Big Data-Analysepipeline definiert und die zusätzliche Herausforderung an die KI- und ML-Fähigkeiten der Nutzer beherrscht werden. Musterarchitekturen, die für Umsetzungen herangezogen werden könnten, existieren keine. Aus diesem Grund verbleibt es eine Herausforderung, Anwendungen von KI und ML für Big Data-Infrastruktur und -Tooling (SMART VORTEX), deren Auswirkungen auf die Nutzerfähigkeiten (EDISON) sowie deren Einbettung in nutzer-befähigende Informationssysteme (IVIS4BigData) zu untersuchen. Um diese Lücke zu schließen hat diese Dissertation mit AI2VIS4BigData ein Referenzmodell entwickelt, das nutzerbefähigende Einbettung von KI und ML in die Big Data-Analysepipeline beschreibt und die KI- und ML-Repräsentation definiert, Nutzerstereotype charakterisiert und eine serviceorientierte Referenzarchitektur bereitstellt. Dieses wurde für sieben Anwendungsdomänen (bspw. Meteorologie) umgesetzt und umfangreich, bspw. über Workshops oder Studien evaluiert
Book review: Electropolishing
No full textThe book „Electropolishing“ by M. Buhlert is dealing with the electrolytic brightening, smoothing and deburring of technical materials like steel, copper, brass, aluminum, titanium and magnesium. The book content covers the basics and the main influencing parameters of the electropolishing process and provides detailed and application orientated procedure information for technical relevant materials.
After a brief introduction into the topic of electrolytic polishing in the first chapter, the author gives a detailed description of the basic electrochemical reaction mechanisms in the second chapter. The reader will also be informed about the physical and chemical parameters, which control the electrochemical removal process. Additionally, the author reports about the advantages and disadvantages of electropolishing.
The third chapter provides a detailed insight into the influencing manufacturing parameters affecting the results of the electropolishing process. Distinct aspects of the manufacturing e.g. workpiece preparation, electrolyte composition, polishing time, electrolyte temperature and alloy compositions will be discussed in detail by the author.
In the fourth and fifth chapter, the author gives useful hints and information about the parameter variation and the suitable handling of hull cells for the optimization of the removal process, followed by a particular overview about material specific electrolyte mixtures for common-used technical metals and alloys, like steel, aluminum, brass, magnesium, copper and titanium.
The sixth chapter of the book provides many selected manufacturing results and investigations on electropolishing of different metals and alloys, which allows the reader the opportunity to develop a systematic understanding of the topic and to adopt the knowledge on the optimization of his own electropolishing process.
Finally, the book concludes with a brief chapter about some remarks with respect to the work safety and environmental efforts.
In summary, this book contains a very detailed and clear arranged overview about the electropolishing method for the surface optimization process. For this reason, it is a suitable and useful lecture for people, which want to take an in depth look into the topic in order to start using this method or are interested in optimize their existing electropolishing processes
A Visual Analytics Technique to Compare the Performance of Predictive Models
No full textDecisions that people make every day are affected by the information available in a given moment. Predictive models are used to estimate future values. For a given set of data and an analysis goal, the results of the models can vary, so it is important to select the most accurate model for the set of data. This paper proposes a Visual Analytics technique for comparing the performance of predictive models. It consists of four main components that support the tasks of the Keim’s Visual Analytics Mantra: “analyze first, show the important, zoom, filter and analyze further, details on demand”. The first component, analyze data, by building predictive models using various machine learning algorithms; the other three components are interactive visualizations that show the important results found by the models, zoom and filter on results of interest and finally, further analyze the selected results by showing details on the data
AI2VIS4BigData: Qualitative Evaluation of an AI-Based Big Data Analysis and Visualization Reference Model
No full text
Genealogia rodziny Jędryckich z Biskupca
No full textIn 1986. Janusz Jasinski gave print of diary of Franiszek Lieder under the title „Warmia my young years”, in which, among others - in connection with the work of his father Jan Lieder as a teacher in Biesowo - appears silhouette of his successor in the local school, Józef Jędrycki, derived from Biskupiec. The figure of Jędrycki gave rise to the presentation of relationships of Jędrycki family from Biskupiec (persons with this name we meet in other cities of Warmia). After years - through access to Warmia vital records, brought in 2002 from Regensburg to the Archives of the Archdiocese of Warmia in Olsztyn - you can confront the arguments of the diarist and publisher with the genealogical information contained in these valuable sources. An important data complement to the article topic from the oldest surviving book of baptisms and the oldest books of the dead from parish Biesowo stored in Sächsisches Staatsarchiv. Staatsarchiv Leipzig (this information was obtained thanks to the kindness and courtesy of the Leipzig archivist Dr. Thoralf Handke, whom the author would like to thank for understanding and the effort to carry out the query source), and the oldest book of the parish Biesowo marriages and other vital records, stored in the Archives of the Archdiocese of Warmia in Olsztyn
Towards a User-Empowering Architecture for Trustability Analytics
No full textMachine learning (ML) thrives on big data like huge data sets and streams from IOT devices. Those technologies are becoming increasingly commonplace in our day to day existence. Learning autonomous intelligent actors (AIAs) impact our lives already in the form of, e.g. chat bots, medical expert systems, and facial recognition systems. Doubts concerning ethical, legal, and social implications of such AIAs become increasingly compelling in consequence. Our society now finds itself confronted with decisive questions: Should we trust AI? Is it fair, transparent, and respecting privacy? An individual psychological threshold for cooperation with AIAs has been postulated. In Shaefer’s words: “No trust, no use”. On the other hand, ignorance of an AIA’s weak points and idiosyncrasies can lead to overreliance. This paper proposes a prototypical microservice architecture for trustability analytics. Its architecture shall introduce self-awareness concerning trustability into the AI2VIS4BigData reference model for big data analysis and visualization by borrowing the concept of a “looking-glass self” from psychology
Towards a Conceptual Modeling of Trustworthiness in AI-Based Big Data Analysis
No full textThis paper introduces a conceptual mathematical framework for evaluating trustworthiness in AI-based big data analysis, emphasizing the critical role trust plays in the adoption and effectiveness of AI systems. The proposed model leverages the trustworthy AI-based big data management (TAI-BDM) reference model, focusing on three fundamental dimensions of trustworthiness: validity, capability, and reproducibility. It formalizes these dimensions mathematically, embedding them into a unified three-dimensional state space that enables the quantification of trustworthiness throughout AI-supported data exploration processes. It further defines update functions capturing the impact of individual data manipulation steps on overall system trustworthiness. Additionally, the paper proposes a scalar metric to integrate and evaluate these dimensions collectively, providing a practical measure of the overall trustworthiness of the system. The paper presents a starting point for modeling trustworthiness in TAI-BDM applications
An Information System Supporting Insurance Use Cases by Automated Anomaly Detection
No full textThe increasing availability of vast quantities of data from various sources significantly impacts the insurance industry, although this industry has always been data driven. It accelerates manual processes and enables new products or business models. On the other hand, it also burdens insurance analysts and other users that need to cope with this development parallel to other global changes. A novel information system (IS) for artificial intelligence (AI)-supported big data analysis, introduced within this paper, shall help to overcome user overload and to empower human data analysts in the insurance industry. The IS research’s focus lies neither in novel algorithms nor datasets but in concepts that combine AI and big data analysis for synergies, such as usability enhancements. For this purpose, this paper systematically designs and implements an AI2VIS4BigData reference model to help information systems conform to automatically detect anomalies and increase its users’ confidence and efficiency. Practical relevance is assured by an interview with an insurance analyst to verify the demand for the developed system and derive all requirements from two insurance industry user stories. A core contribution is the introduction of the IS. Another significant contribution is an extension of the AI2VIS4BigData service-based architecture and user interface (UI) concept on AI and machine learning (ML)-based user empowerment and data transformation. The implemented prototype was applied to synthetic data to enable the evaluation of the system. The quantitative and qualitative evaluations confirm the system’s usability and applicability to the insurance domain yet reveal the need for improvements toward bigger quantities of data and further evaluations with a more extensive user group
