Digital Library of Gesellschaft für Informatik e.V.
Not a member yet
40758 research outputs found
Sort by
Bereitstellung und Präsentation der Analysedaten zu Verdachtsfällen auf PSM-Vergiftung von Honigbienen in einem Multiakteur-Softwaresystem
No full textIm Projekt Sen2Bee wird ein Softwaresystem zur Unterstützung der Analyse von Bienenvergiftungen entwickelt. Es integriert und visualisiert relevante Informationen wie Feldgeometrien, Kulturen, Phänologie und meteorologische Bedingungen. Die Akteure umfassen Imkernde, die Vorfälle melden, Amtspersonen, die Proben entnehmen, und Fachleute der UBieV, welche die Untersuchungen durchführen. Die Systementwicklung adressiert Herausforderungen wie Datenintegration, Interoperabilität und Benutzerfreundlichkeit. Für eine Reihe dieser Herausforderungen wurden Lösungen vorgeschlagen und diskutiert. Darunter die Auswahl der passenden Quelle für Wetterdaten zur qualitativen Beurteilung der Bienenflugwahrscheinlichkeit, Einbindung der Daten aus der JKI-Infrastruktur, Optimierungsstrategien für die Antwortzeiten der WCPS-Abfragen sowie Entwicklung einer App für bessere Benutzerakzeptanz
Achilles' SPEar: Using Metamorphic Testing to Find Bugs in Stream Processing Engines
No full textStream Processing Engines (SPEs) are critical for real-time data processing, relying on aggressive optimizations to meet performance demands. Ensuring the reliability of such systems requires robust testing, yet testing remains costly and challenging due to the oracle problem. This paper investigates adapting query partitioning, a metamorphic testing technique, to the domain of SPEs. We present Achilles, an automated testing framework for SPEs. Achilles utilizes query partitioning to automatically generate, execute, and evaluate diverse test cases, reducing manual effort and targeting stream operators like filters and windowed aggregations. Our evaluation highlights Achilles' ability to detect unique bugs, including division-by-zero errors and predicate evaluation flaws. We further analyze framework parameters and demonstrate the effectiveness of predicate-based query partitioning, finding that increasing predicate depth boosts bug detection by 10%
Improved Bounds for Rounding Errors in Quantum Circuit Simulators
No full textSimulators play a crucial role in the development of quantum software, yet they differ from actual quantum devices in that their computations are carried out using floating-point arithmetic rather than real arithmetic. In [7], we introduced bounds on the errors that may arise due to these discrepancies. In the present work, we extend and refine these bounds, demonstrating that we can achieve tighter estimates that scale more efficiently with the number of gates in a quantum circuit. Furthermore, the newly derived bounds can be computed with minimal relaxation, making them practically applicable. We show that these improved bounds are effective in excluding significant errors across a wide range of quantum circuits, thus providing a more reliable framework for simulating quantum systems
Federated Learning for data-sovereign training of weed and crop detection machine learning models
No full textAs digitalization progresses, AI applications are increasingly finding their way into various agricultural settings. These solutions have great potential, but the underlying AI models require a lot of diverse and realistic data for accurate predictions. This could be realized by centrally collecting data from many farms and attempting to build highly generalized models from them. However, centralized data collection raises privacy concerns for farmers since they want to protect their company-specific information. Federated Learning (FL) offers a solution by training models locally and sharing only model parameters, preserving data sovereignty during training. This paper presents experiments on weed and crop detection systems, demonstrating the potential of FL in agriculture. We first reproduce experiments from literature using YOLOv8 to benchmark against traditional centralized models. We then apply FedAvg to compare FL with centralized training in homogenous and heterogenous cases. Promising results show FL maintains performance while safeguarding data privacy, even across different datasets and locations. The current limitations of our approach will be discussed, as well as aspects of future work
Fast Compression of Floating-Point Values with Exponent/Mantissa Shuffling
No full textWe investigate fast compression algorithms for vectors of floating-point values, as these frequently appear in numerical computations. Our proposal bases on Snappy yet uses a modified data layout. Use of data chunks allows parallel thread-based compression without loss of compression quality. Our proposal beats previous proposals and standard compression algorithms in compression ratio and/or speed. The gain in compression ratio and speed can be used to shorten communication phases, e.g., between CPUs and accelerators
Hyperspectral band selection using segmented autoencoders for visual quality assessment of food products
No full textIn the food industry, natural products that fail to meet quality standards must be removed during processing. To automate this resource-intensive and error-prone process, this work presents a system for AI-supported quality detection of natural products based on hyperspectral images. The implemented system consists of three components: background filter, dimension reduction using autoencoders, and CNN classifier for quality assessment. By training multiple autoencoders for different spectral segments, the proposed architecture can extract the essential spectral information from a given input image, selecting the most informative spectral bands. The system was evaluated on datasets containing chicken legs and potatoes, recorded by a hyperspectral sensor with 224 spectral bands. The results show that the system enables efficient processing of relatively large hyperspectral datasets. Furthermore, the dimension reduction carried out is suitable for the robust classification of defective natural products
Computational Capabilities and Compiler Development for Neutral Atom Quantum Processors --- Connecting Tool Developers and Hardware Experts
No full textGraph-based QSS: A Graph-based Approach to Quantifying Semantic Similarity for Automated Linear SQL Grading
No full textDetermining the Quantified Semantic Similarity (QSS) between database queries is a critical challenge with broad applications, from query log analysis to automated SQL skill assessment. Traditional methods often rely solely on syntactic comparisons or are limited to checking for semantic equivalence. This paper introduces Graph-based QSS, a novel graph-based approach to measure the semantic dissimilarity between SQL queries. Queries are represented as nodes in an implicit graph, while the transitions between nodes are called edits, which are weighted by semantic dissimilarity. We employ shortest path algorithms to identify the lowest-cost edit sequence between two given queries, thereby defining a quantifiable measure of semantic distance. An empirical study of our prototype suggests that our method provides more accurate and comprehensible grading compared to existing techniques. Moreover, the results indicate that our approach comes close to the quality of manual grading, making it a robust tool for diverse database query comparison tasks
Code Generation for Niche Programming Languages with Large Language Models
No full textCode generation is a prominent use-case for Large Language Models (LLMs). Specialized LLMs such as CodeLlama or Codestral are trained on a large variety of programming languages and achieve a strong performance on coding tasks. However, when applied to less common programming languages which are not included in their pre-training corpus, their performance decreases. In this work, we describe an approach to integrate a LLM in the context of a coding copilot for specific applications where code shall be generated in a niche general-purpose programming language. We study the use of an intermediate domain-specific-language to limit the scope to the application-specific needs, and to enable the LLM to reliably generate code in such an application-specific scenario. We evaluate this method on two use-cases: Generating constraints in the context of product configuration using the MiniZinc constraint language, and generating test specifications in the context of railway infrastructure using the Balise Telegram Test Language. Our results show that defining an intermediate scope-limited DSL improves the performance of an LLM in our evaluated application-specific code generation scenarios. However, we can not guarantee that the presented performance results are generalizable to all scenarios
Improving Unnesting of Complex Queries
No full textSQL allows for very flexible nesting of queries, including subqueries that access attributes by the outer part of the query. These correlated subqueries simplify query formulation, but their execution is very inefficient, leading to O(n^2) runtime complexity. Which can become prohibitively expensive for large databases. Query optimizers therefore try to unnest, i.e., decorrelate the dependent queries. Existing decorrelation techniques, however are either limited in scope or lead to suboptimal execution plans when correlated queries are stacked repeatedly inside each other. In this work we present a generalized unnesting approach that can handle deep nestings of correlated subqueries and that generalizes to complex query constructs, including recursive SQL. This generalized unnesting improves the asymptotic complexity, and thus can lead to dramatic performance improvements in the affected queries