Machine-interpretable dataset and service descriptions for heterogeneous data access and retrieval

Abstract

To refer to or to cite this work, please use the citation to the published version: Dimou, A., Verborgh, R., Vander Sande, M., Mannens, E., and Van ABSTRACTThe rdf data model allows the description of domain-level knowledge that is understandable by both humans and machines. rdf data can be derived from different source formats and diverse access points, ranging from databases or files in csv format to data retrieved from Web apis in json, Web Services in xml or any other speciality formats. To this end, vocabularies such as rml were introduced to uniformly define how data in multiple heterogeneous sources is mapped to the rdf data model, independently of their original format. This approach results in mapping definitions that are machine-processable and interoperable. However, the way in which this data is accessed and retrieved still remains hard-coded, as corresponding descriptions are often not available or not taken into account. In this paper, we introduce an approach that takes advantage of widelyaccepted vocabularies, originally used to advertise services or datasets, such as Hydra or dcat, to define how to access Web-based or other data sources. Consequently, the generation of rdf representations is facilitated, as the description of the interaction models with the original data remains independent, interoperable and granular.

Conformance Test Cases for the RDF Mapping Language (RML)

“…RML [9] is defined as an extension of R2RML to specify rules for generating knowledge graphs from data in different formats, such as CSV, JSON, XML, and different access interfaces, e.g., open data connectivity and Web APIs [14]. Different other languages build upon RML for generating knowledge graphs from heterogeneous data sources, e.g., xR2RML [10] or RMLC [15].…”

“…Each of these specifications has several related tools that support them. A set of test cases was defined for each one of them (SPARQL test cases 12 , RDF 1.1 test cases 13 , SHACL test cases 14 , and R2RML and Direct Mapping test cases 15 , respectively) that provides useful information to choose the tool that fits better to certain needs. It is also a relevant step in the standardisation process of an technology or specification.…”

ILastic: Linked Data Generation Workflow and User Interface for iMinds Scholarly Data

“…Even though OpenAIRE LOD [23], the Open Access Infrastructure for Research in Europe 13 , was recently launched, it still relied on a custom solution to generate its Linked Data set from OpenAIRE Information Space which cannot be reused by any other data publisher. As performance and scalability were major concerns, a MapReduce [7] processing strategy was preferred.…”

“…It uses D2R conversions, as DBLP, but it also requires data derived from different data sources to be turned into CSV files which, on their turn, are ingested into a SQL database. With our proposed approach, we manage even to avoid this preprocessing step and directly use the original data sources [13], reducing the required effort and maintenance costs and increasing at the same time the reusability of our workflow and the reproducibility of the generated Linked Data.…”

“…It considers the input data as it is retrieved from the DataTank and aims to directly semantically annotate them with the aforementioned vocabularies and ontologies. The rmlprocessor relies on machine-interpretable descriptions of the data sources [13]. To access the iMinds data warehouse, the rmlprocessor relies on the api's description which is defined using the Hydra vocabulary 38 .…”