Logo des Repositoriums

Auflistung nach Schlagwort "Knowledge Graph"

1 - 10 von 10
  • Konferenzbeitrag
    Automated Reasoning for Conflict Solving in Knowledge Graphs
    (INFORMATIK 2024, 2024) Fähndrich, Johannes; Wischow, Maik
    Forensic application of Methods of AI depends on the level of trust towards automated reasoning. Automated reasoning leads necessarily to conflicts, and with that to the need for adaptation. Knowledge Graphs are an existential part of formalization in complex systems, e.g. as representation of beliefs of an AI. Strong AI, and with that one of the two main research areas of the early 21st century in Computer Science, struggles with the representation of conflicting beliefs, as well as with strategies for their resolution. We present a template based approach with an implementation on detecting and resolving conflicts in belief systems leading to a deeper insight into AI and its ability of self reflection. Without the understanding of how beliefs are handled in strong AI systems, the application to forensics is hurdled.
  • Konferenzbeitrag
    CLOCQ: A Toolkit for Fast and Easy Access to Knowledge Bases
    (BTW 2023, 2023) Christmann, Philipp; Roy, Rishiraj Saha; Weikum, Gerhard
    Curated knowledge bases (KBs) store vast amounts of factual world knowledge, and are therefore ubiquitous in many information retrieval (IR) and natural language processing (NLP) applications like question answering, named entity disambiguation, or knowledge exploration. Despite that, accessing information from complete knowledge bases is often a daunting task. Researchers and practitioners typically have crisp use cases in mind, for which standard querying interfaces can be overly complex and inefficient. We aim to bridge this gap, and release a public toolkit that provides functionalities for common KB access use cases, and make it available via a public API. Experiments show efficiency improvements over existing KB interfaces for various important functionalities.
  • Konferenzbeitrag
    Domain-agnostic Intelligent Digital Twins: Merging of Application-near Knowledge Representations with the Proactive Internet of Digital Twins (IoDT)
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Lehmann, Joel; Lober, Andreas; Häußermann, Tim; Rache, Alessa; Baumgärtel, Hartwig; Reichwald, Julian
    Flexibility and reconfigurability regarding mass-customized products are the key premises of future industrialization. The amalgamation of Digital Twin (DT) approaches, knowledge-representing technologies, and skill-based engineering methodologies allow such approaches to emerge. This paper deals with the integration of knowledge representations as a driver for holistic system intelligence ranging from physical assets to the Internet of Digital Twins (IoDT) within the Digital Twin Reference Model (DTRM). This is what it takes to spur DTs to intelligence and proactive collaboration behaviors. Despite the domain-agnostic, universal applicability of the concept, exemplary implementation approaches from the field of production and intralogistics illustrate the feasibility. Specific domain knowledge can be dynamically aggregated and used as a basis for negotiation scenarios between DTs.
  • Konferenzbeitrag
    Knowledge Graph Processing Made (more) Simple
    (40 Years EMISA 2019, 2020) Lausen, Georg
    Knowledge graphs based on RDF and SPARQL are gaining popularity for integrated semantic representation of structured and unstructured data. As knowledge graphs in practical applications tend to become huge, distributed processing using Apache Spark SQL and Hadoop on top of a compute cluster is attractive. For the corresponding relational representation of a knowledge graph, a simple relational design using only one single table is proposed. Consequently no time consuming relational design considerations are required and newly discovered RDF data can be integrated with nearly no extra additional relational design effort.
  • Textdokument
    Konzept und Prototyp einer dezentralen Wissensinfrastruktur zu Hochschuldaten für Mensch und Maschine
    (INFORMATIK 2017, 2017) Meister, Vera; Jetschni, Jonas; Kreideweiß, Sebastian
    Der Beitrag beschreibt den Stand der Entwicklungen für eine dezentrale Wissensinfrastruktur zu Hochschuldaten, welche Mehrwertdienste für Mensch und Maschine unterstützt. Im Fokus stehen zunächst wenig volatile Daten zu Studiengängen, die aktuell mit hohem Aufwand in den verschiedensten technischen und organisationalen Strukturen vorgehalten werden. Das zeigt eine aktuelle Analyse der Ausgangslage. Das Architekturkonzept kann als Knowledge Graph beschrieben werden, der Webseiten von Hochschul-Content-Management-Systemen (CMS) als primäre Wissensquellen nutzt. Dies wird zunächst durch CMS-Extensions erreicht, die auf Semantic-Web-Technologien, insbesondere auf schema.org, JSON-LD und SPARQL setzen. Die Anbindung weiterer strukturierter und semi-strukturierter Wissensquellen erfolgt in transparenten Datenintegrationsprozessen, welche individuelle Orientierung ebenso wie anforderungsspezifische Datenaktualisierung unterstützen. Neben der systematischen Darstellung des Architekturkonzeptes werden Meilensteine der prototypischen Implementierung erläutert. Der Beitrag schließt mit einem Ausblick auf Anforderungen und Rahmenbedingungen einer produktiven Implementierung.
  • Konferenzbeitrag
    On which legal regulations is a public service based? Fostering transparency in public administration by using knowledge graphs
    (INFORMATIK 2023 - Designing Futures: Zukünfte gestalten, 2023) Feddoul, Leila; Raupach, Maximilian; Löffler, Felicitas; Babalou, Samira; Hoyer, Jonas; Mauch, Marianne; König-Ries, Birgitta
    Knowledge about digitization in the public administration is complex and scattered. Information about legal regulations, methods, processes, APIs, metadata and data standards, registers, and terminologies are spread across different platforms. Hence, it is difficult for employees, developers, and decision makers to figure out what data standards, formats, and APIs are relevant for the digitization of a specific public service. Training administrative employees and IT companies requires to gather and link the required knowledge first in a well-structured and accessible manner. We address this need for more shared and transparent knowledge in the digitization of public services in Germany. We propose a first version of an ontology (GerPS-onto) for public administration instantiated by one example german public service. We utilize semantic modelling and Linked Data technologies to enable appropriate data and process descriptions that are readable for humans and machines. We also demonstrate how an existing process description of a public service can be linked to existing terminologies and evaluate the resulting ontology using domain-specific competency questions that are translated to SPARQL queries.
  • Konferenzbeitrag
    A Provenance Management Framework for Knowledge Graph Generation in a Web Portal
    (BTW 2023, 2023) Kleinsteuber, Erik; Babalou, Samira; König-Ries, Birgitta
    Knowledge Graphs (KGs) are the semantic backbone for a wide variety of applications in different domains. In recent years, different web portals providing relevant functionalities for managing KGs have been proposed. An important functionality such portals is provenance data management of KG generation process. Capturing, storing, and accessing provenance data efficiently are complex problems. Solutions to these problems vary widely depending on many factors like the computational environment, computational methods, desired provenance granularity, and much more. In this paper, we present one possible solution: a new framework to capture coarse-grained workflow provenance of KGs during creation in a web portal. We capture the necessary information of the KG generation process; store and retrieve the provenance data using standard functionalities of relational databases. Our captured workflow can be rerun over the same or different input source data. With this, the framework can support four different applications of provenance data: (i) reproduce the KG, (ii) create a new KG with an existing workflow, (iii) undo the executed tools and adapt the provenance data accordingly, and (iv) retrieve the provenance data of a KG.
  • Textdokument
    Towards a Semantic Toolbox for Reproducible Knowledge Graph Generation in the Biodiversity Domain - How to Make the Most out of Biodiversity Data
    (INFORMATIK 2021, 2021) Babalou, Samira; Schellenberger Costa, David; Kattge, Jens; Römermann, Christine; König-Ries, Birgitta
    Knowledge Graphs are widely regarded as one of the most promising ways to manage and link information in the age of Big Data. Their broad uptake is still hindered by the large effort required to create and maintain them, though. In this paper, we propose the creation of a semantic toolbox that will support data owners in transforming their databases into reproducible, dynamically extendable knowledge graphs that can be integrated and jointly used. We showcase the need, potential benefits and first steps towards the solution in our example domain, biodiversity research.
  • Konferenz-Abstract
    Using Knowledge Graphs to Detect Enterprise Architecture Smells
    (EMISA 2022, 2022) Hacks, Simon; Smajevic, Muhamed; Bork, Dominik
  • Textdokument
    Using Knowledge Graphs to Manage a Data Lake
    (INFORMATIK 2020, 2021) Dibowski, Henrik; Schmid, Stefan
    Knowledge graphs as fundamental pillar of artificial intelligence are experiencing a strong demand. In contrast to machine learning and deep learning, knowledge graphs do not require large amounts of (training) data and offer a bigger potential for a multitude of domains and problems. This article shows the application of knowledge graphs for the semantic description and management of data in a data lake, which improves the findability and reusability of data, and enables the automatic processing by algorithms. Since knowledge graphs contain both the data as well as its semantically described schema (ontology), they enable novel ontology-driven software architectures, in which the domain knowledge and business logic can completely reside on the knowledge graph level. This article further introduces such a use case: an ontology-driven frontend implementation, which is able to fully adapt itself based on the underlying knowledge graph schema and dynamically render information in the desired manner.