|
Title:
|
TRACEABLE RDF DATA GENERATION FROM GERMAN LEGISLATIVE DOCUMENTS THROUGH KNOWLEDGE GRAPH TRANSFORMATIONS |
|
Author(s):
|
Alexander Gashkov, Oliver Schmidtke and Andreas Both |
|
ISBN:
|
978-989-8704-62 |
|
Editors:
|
Paula Miranda and Pedro Isaías |
|
Year:
|
2024 |
|
Edition:
|
Single |
|
Keywords:
|
Knowledge Graphs, RDF Data Transformation, Semantic Enrichment, Legislative Documents, German Federal Law |
|
Type:
|
Full |
|
First Page:
|
259 |
|
Last Page:
|
266 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
The current shift toward a data-driven and web-based society is being driven by the data available. However, it is not just
the availability but also the traceability, e.g., regarding the structure of documents and their semantics.
In particular, because of the permanent growth and changing of normative documents, transforming them into a
machine-readable and machine-interpretable form is crucial for trustworthy data-driven applications.
However, often the provided data formats (e.g., XML) have several flaws, in particular, not providing the semantics of
the data or limited metadata. For instance, access to German laws is provided via a web portal as text and PDF files (for
humans) and as an XML file (a machine-readable format). However, an XML file downloaded from the site features
limited semantics, e.g., it does not possess cross-references between laws or paragraphs.
This paper presents an RDF-driven, step-by-step, extensible transformation process in which each transformation step
leads to a knowledge graph. Hence, data transformation steps can be validated, s.t., the validity and trustworthiness of the
generated data are increased and - in the best case - can be guaranteed. As a use case, we generate a knowledge graph
from the German federal laws. The transformation steps include cleaning up the data, adding types, and establishing links
between laws and their paragraphs.
The contributions of this paper are a process, the dataset representing the RDF data of the German federal law, and
additional insights for the particular use case. |
|
|
|
|
|
|
