EUVIMEDEUVIMED entdecken
Crossref

From Open Information Extraction to Probabilistic Fusion: Semantic Retrieval Pipelines for Enterprise Knowledge Graph Construction

Sriram Ghanta

International Journal of Research and Applied Innovations · 20.05.2023 · 06 (03)

DOI: 10.15662/ijrai.2025.0802010 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Research and Applied Innovations, online erschienen am 20.05.2023. Verfasst von Sriram Ghanta.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The exponential growth of unstructured enterprise data spanning documents, logs, emails, reports, and web content has intensified the demand for scalable mechanisms capable of extracting, organizing, and retrieving knowledge in machine-interpretable forms. Knowledge Graphs (KGs) have consequently emerged as a foundational representation for modeling entities, relationships, and contextual semantics across heterogeneous and distributed information sources, enabling more advanced analytics, reasoning, and decision support. This article presents a systematic exploration of semantic retrieval pipelines for enterprise knowledge graph construction, tracing their evolution from early Open Information Extraction (OpenIE) systems to more sophisticated probabilistic knowledge fusion architectures. Drawing upon seminal systems such as TextRunner and SigmaKB, we analyze how successive pipeline stages including large-scale text ingestion, relation extraction, semantic filtering, entity normalization and disambiguation, and probabilistic knowledge fusion work in concert to transform noisy, unstructured data into coherent and reliable enterprise knowledge graphs. The discussion synthesizes recurring architectural patterns observed across foundational systems, examines practical challenges encountered in large-scale enterprise deployments such as noise management, ambiguity resolution, scalability, and trust and highlights emerging directions toward AI-augmented semantic retrieval, where machine learning and neural representations increasingly complement symbolic knowledge representations to enhance robustness, adaptability, and semantic depth

Zitieren

APA

Ghanta, S. (2023). From Open Information Extraction to Probabilistic Fusion: Semantic Retrieval Pipelines for Enterprise Knowledge Graph Construction. International Journal of Research and Applied Innovations, 06(03). https://doi.org/10.15662/ijrai.2025.0802010

Vancouver

Ghanta S. From Open Information Extraction to Probabilistic Fusion: Semantic Retrieval Pipelines for Enterprise Knowledge Graph Construction. International Journal of Research and Applied Innovations. 2023;06(03). doi:10.15662/ijrai.2025.0802010

BibTeX
@article{Ghanta2023,
  author = {Ghanta, Sriram},
  title = {From Open Information Extraction to Probabilistic Fusion: Semantic Retrieval Pipelines for Enterprise Knowledge Graph Construction},
  journal = {International Journal of Research and Applied Innovations},
  year = {2023},
  volume = {06},
  number = {03},
  doi = {10.15662/ijrai.2025.0802010}
}
Kostenlos starten →