9-11 October 2018
Lisbon
Europe/Lisbon timezone

B2FIND - Searching for Research Data across Disciplines

Not scheduled
15m
Lisbon

Lisbon

ISCTE, University of Lisbon
Poster Posters

Speaker

Claudia Martens (German Climate Computing Center [DKRZ])

Description

Searching for relevant information is the beginning of every research project. Finding appropriate data that encourage scientific work is not an easy task in an increasingly interconnected research world but it´s the foundation where B2FIND is attached. Its greatest value is the possibility to search and find research data across scientific disciplines and research areas. B2FIND is not meant to be a specialized search portal for a specific community. But it may enable an enhanced view of which data from what community could be useful for a specific topic hence it could accelerate knowledge by revealing unexpected connections. The backbone of B2FIND is its ingestion process that consists of three steps. First the metadata records - provided by various research communities - are harvested. Afterwards the raw metadata records are converted and mapped to unified key-value dictionaries as specified by the B2FIND schema. The main challenge here is the divergence in metadata standards and schemes; these differences correspond to the specific communities´ needs. Nevertheless they impede the extraction of content related information to enable a cross-disciplinary search. Thus the semantic mapping of the non-uniform, community specific metadata to homogenous structured datasets is the most subtle and challenging task. To assure and improve metadata quality this mapping process is accompanied by • iterative and intense exchange with community representatives, • usage of controlled vocabularies and community specific ontologies and • formal and semantic mapping and validation. Finally the mapped and checked records are uploaded as datasets to the catalogue which is based on CKAN, an open source data portal software that provides a rich RESTful JSON API and uses SOLR for indexing. The homogenization of community specific data models and vocabularies enables not only a unique presentation of these datasets as tables of field-value pairs but also an interdisciplinary and cross-community search with geospatial and temporal search functionalities. Results from a free text search may be narrowed down by using the facets. B2FIND offers support for new communities interested in publishing their data within EUDAT and EOSC.

Summary

The metadata service B2FIND plays a central role within the pan-European collaborative research data infrastructure EUDAT-CDI by providing a simple and user-friendly discovery portal to find and access research data collections stored in EUDAT data centers as well as in other repositories. Therefore metadata collected from heterogeneous sources are stored in a comprehensive joint metadata catalogue and made searchable via an open data portal. Furthermore B2FIND provides transparent access to the scientific data objects through the given references and identifiers in the metadata – thus supporting the first two pillows of FAIR data principles

Type of abstract Poster

Primary author

Claudia Martens (German Climate Computing Center [DKRZ])

Co-author

Heiner Widmann (German Climate Computing Center [DKRZ])

Presentation Materials

There are no materials yet.