Task area

Task Area 3: Data Services

TA3 aims to provide a set of robust and reliable key services and link them with further data collections and tools in an interoperable “4Memory Data Space” via a knowledge graph and standardised technical interfaces. It will offer a technical counselling service to empower the 4Memory community to expand the Data Space with their own services. TA3 pursues five goals: (1) operating a set of key services in a reliable and sustainable manner, including those for storage and long-term archiving of research data for the historically oriented disciplines; (2) implementing a knowledge graph based on existing metadata as a representation of the new digital knowledge order; (3) making the existing repositories, DH tools and data offerings of memory organisations relevant to historically oriented disciplines interoperable and searchable by defining and setting-up a 4Memory Data Space; (4) combining data spheres of memory institutions and research projects in a virtual indexing layer with cooperative data enrichment; (5) ensuring the FAIRness of research data in 4memory.

If you do not want to miss any activities of TA Data Connectivity, sign in for our mailing list: https://lists.nfdi.de/postorius/lists/nfdi4memory-ta3-community.lists.nfdi.de/

Measures

Measure 1: Standardised and interoperable APIsPlease click button to expand or collapse content

Historians use heterogeneous data sources. The compilation of project-specific resource collections spanning multiple data sources is therefore a laborious intellectual process. Unified programming interfaces allow standardised, interoperable access to these heterogeneous data. M1 will therefore define such an interoperable API together with the providers of key data sources and foster their implementation. This is important preparatory work for the Data Space planned in M3. We will closely monitor international developments such as IIIF, SPARQL or GraphQL and coordinate with the other consortia of the MoU group as well as the “Common Infrastructure” section of the NFDI.

Measure 2: Technical implementation of a 4Memory Knowledge Graph Please click button to expand or collapse content

M2 builds the KG’s technical foundation alongside a distributed editorial system enabling federated maintenance: various projects will be able to participate and contribute. Among other benefits, distributed editing gives non-European communities decisive influence over the representation of their cultures in knowledge systems, supporting the observance of CARE principles. The federated approach will raise complex questions of data quality that are addressed in TA1. The KG’s core component is an ontology conceptualised in TA2 in line with international standards. It will cover specific concepts and knowledge domains relevant to historians, such as temporal fuzziness, the changing semantics of terms over time, hypotheses, and temporal and spatial networks. The KG enables the semantic linking of the heterogeneous data sources of the Data Space (M3), allowing for interlinking with external resources, e.g. from other NFDI consortia or internationally. All the KG’s data will be queryable via a SPARQL endpoint.

Measure 3: Creation of a 4Memory Data Space Please click button to expand or collapse content

M3 designs and implements the 4Memory Data Space, a federation of heterogeneous data sources from research and memory organisations, which we connect via common interfaces (M1) and a Knowledge Graph (M2). This allows researchers to perform cross-searches and compile corpora from different sources.

Measure 4: Technical counselling service on interoperability and reusability of dataPlease click button to expand or collapse content

To continuously expand 4Memory’s emerging ecosystem of services while increasing the interoperability of services and reusability of data, TA3, with the other TAs, establishes a counselling service to support researchers and project leaders in applying for and carrying out projects and to advise them on selecting suitable services and data for their research questions. The service will consider standards identified or developed in TA2 and become a part of the 4Memory helpdesk organised by TA6-M4.

Measure 5: 4Memory Data LabPlease click button to expand or collapse content

The 4Memory Data Lab provides a technical infrastructure for the evaluation and deployment of innovative technologies and applications for annotation, feature detection and data analysis as well as text recognition. It will combine seamless access to relevant data sets and corpora via the Data Space with access to a high-performance hardware cluster. Suitable applications and technologies based on machine learning algorithms are developed or defined in the Method Lab (TA5-M2), including support for researchers on how to employ these emerging methodologies. The resulting annotations and deep indexing will flow back to the data-providing institutions via the virtual indexing layer and be available to third parties. In the first funding phase, the technical design is defined and implemented along practical requirements. Subsequently, the Data Lab will be put into operation. In future, a distributed approach with further clusters and corresponding operators and institutions will be pursued.

Measure 6: Design and development of a virtual indexing layerPlease click button to expand or collapse content

TA3 works with research projects, memory organisations and citizen scientists to design and implement a virtual indexing layer, combining memory institutions’ indexing information (source descriptions) and deep indexing from projects (content annotation). The virtual layer enables the combined data’s comprehensive search and retrieval, fostering reuse of valuable indexing information from projects. It will enable cooperative indexing or enrichment of existing indexing data that can be reintegrated into the original systems. We aim for easy transfer of annotations and metadata generated within the 4Memory Data Lab (M5). In the first funding period, TA3 develops and coordinates a concept with the community: implementation occurs in the second funding phase.

Task Area 2: Data Connectivity

TA2 advances the linking and contextualising of research data in the historical domain to enable a coherent historical record in the digital age and provide the foundation for the planned 4Memory Data Space. It lays the ground for historically sensitive, unambiguous data descriptions; the integration of research-led historical categorisation and cataloguing by archives, libraries and museums; and contextualisation within larger data structures. Data connectivity is vital to achieving the LINKAGE objectives. Developing and improving authority files, taxonomies, metadata, ontologies and APIs will increase the linking of research, memory institutions and infrastructures via a knowledge graph, also adding to the representation of a knowledge order for the digital future of the past. Connecting sources, publication and research data advances the analogue / digital interface of historical source material and data, as both types of material will be accessible in one data space. Open, jointly developed authority files generate standards for historical research data and sustainability.

If you want to reach out to us, please use this mail address: ta2-nfdi4memory@geschichte.uni-halle.de

To learn more about the work of and on Data Connectivity, visit the blog: https://blogs.urz.uni-halle.de/nfdi4memory/

If you do not want to miss any activities of TA Data Connectivity, sign in for our mailing list: https://lists.nfdi.de/postorius/lists/nfdi4memory-ta2.lists.nfdi.de/

Services

R:hovono - register for historical and object-related vocabularies and standard dataPlease click button to expand or collapse content

What is "R:hovono"?

The acronym R:hovono stands for the “Register for historical and object-related vocabularies and standard data”, a service that is intended to provide an overview of existing vocabularies, standard data, classifications, thesauri and ontologies to all the disciplines in the humanities and history. The register will comprehensively expand existing structures of information and thus facilitate the findability and categorisation of across-institutions. We are focusing on specific specialised information ranging from metadata, usage scenarios, access options, cross-concordances to technical details. Registering vocabularies is possible either via LimeSurvey or directly in the FactGrid database. The service aims to facilitate the use uses of vocabularies in the wide range of applications from linking and indexing data to data curation, information enrichment, information extraction and data analysis. Contributions are welcome from the entire field of historical disciplines. The editorial team of TA “Data Connectivity” will accompany the process with advice and with the curation of incoming entries.

The documentation on the register of historical and object-orientated vocabularies and authority data and vocabularies (R:hovono) is published on Zenodo.
The presentation of R:hovono is published on Zenodo.

Measures

Measure 1: Development framework for historical authority data Please click button to expand or collapse content

M1 analyses existing authority data approaches for the long-term development of graduated concepts for search strategies, data curation, data exploitation and quality requirements. This applies both to the subject-specific context of 4Memory and to interdisciplinary and international initiatives. In doing so, 4Memory takes into account authority data that contextualise historical entities in a long-term perspective. TA2 begins by working with four use cases (UC) selected through community meetings and problem stories, bringing together diverse actors from 4Memory’s international network. They cover concepts in different historical periods (UC1 Premodern, UC2 Modern/Contemporary History), historical geographical categories (UC3 Space) and personal relationships/activities and occupations (UC4 Activities). To this end, existing approaches are gathered and screened, analysed and modelled, as well as represented and described in a directory of subject-specific knowledge organisation systems, such as authority data and controlled vocabularies. At the same time, existing activities are networked more closely in order to develop common goals and guidelines in the field of authority data production, this applies to specific concepts (UC 4), especially for a more in-depth cooperation with NFDI4Objects and Berd@NFDI.

Measure 2: Working Group on integrating historical data into authority data Please click button to expand or collapse content

The international working group aims to integrate and link authority data to the GND and other standards. The concepts derive from selected subject-specific authority data provided as use cases from the participants, in order to meet the needs of both compact, largely standardised keywording and diversified processes of scholarly analysis with basically the same networked approaches and concepts. The aim is establishing a two-tiered, linked network of persistently addressable authority data (GND, metadata, search) in relation to dynamically changing data (taxonomies, ontologies, context data) geared towards scholarly analysis. This approach addresses an essential goal for producing FAIR data by maintaining data transparency, traceability and contextualisation from the (analogue or digital) source to the national/international authority data set: it is essential to integrating historical source criticism (contextualisation), developing standards and networking different process levels of RDM (i.e. data production, analysis, provision and reuse).

Measure 3: Data curation and quality measurement interfaces and tools Please click button to expand or collapse content

M3 addresses the conception, planning and preparation/development of tools for data curation, content quality measurement and the automated indexing or enrichment of research data (e.g. the definition of knowledge graphs, preparation for the Semantic Web or contextualisation of information) and providing the groundwork for their realisation in accompanying projects and/or the second NFDI funding phase. Together with TA1, criteria of a graded quality model for the authority data concepts of specific historical sub-areas are developed. Basic concepts for the normalisation and contextualisation of data as well as for its taxonomic preparation are subjected to quality control and prepared for transfer into tools and interfaces (TA1, TA3, TA4, TA5).

Measure 4: Metadata cataloguing and documentation of research data Please click button to expand or collapse content

After evaluating the status quo and its desiderata, expert groups develop common standards and rules to describe and document research data, easing data linkage and aggregation in the 4Memory community and its Data Space, as well as with international services. Scholars will have a shared set of descriptors, authority data and taxonomies to search for sources and research data. Service providers such as archives, libraries and museums will be given clear guidelines on cataloguing and documenting research data based on common standards and rules adapted to them. The inclusion of a defined set of research data – related to publications and as publications in their own right – into the German Historical Bibliography (GHB) and possibly other bibliographies provides a proof of concept for recommended formal descriptors, authority data and taxonomies. As described in the RDM-Strategy, 4Memory’s research community will benefit from the extended GHB by being able to use this new resource in an established environment with direct-access options.

Measure 5: Interoperable data linkage for the Data Space Please click button to expand or collapse content

For the exchange, integration, delivery and reuse of research data, a 4Memory ontology and a metadata schema must be developed by a metadata competence unit. This will allow linking the research data of the 4Memory community within our Data Space and connecting it to other aggregation services such as BASE and Google Dataset Search or future services in NFDI and EOSC. The developed data model must also integrate non-German languages as well as knowledge and organisation systems in order to represent research data from fields such as area studies, to introduce postcolonial perspectives and to secure the international connectivity and visibility of data curated and listed by 4Memory, thus enabling a truly transnational historiography. Additionally, persistent identifiers will allow addressing granular data, improving the citation and reuse of research data through, e.g. linking to micro-publications such as annotations or referencing concepts within taxonomies.

Task Area 1: Data Quality

TA1 assesses and advances the definition, measurement and implementation of research data quality in the historically oriented humanities, enabling further academic study and computational research through FAIR data. It addresses several LINKAGE objectives. Working groups and events Link research, memory institutions and infrastructures. Elaborating methods for data quality Integrates historical source criticism into data services. Assessing and implementing data quality practices strengthens the Network of historically oriented research communities. Quality assured data shape the Knowledge order for the digital future of the past, and quality criteria (e.g. provenance and processing information) Advance the analogue / digital interface of historical source material and data. Data-quality guidance Generates standards for historical research data and advancing sustainability. TA1 deals with content-related, administrative (e.g. legal documentation) and technical aspects of data quality: standards for research data in the historically oriented humanities are (further) developed and implemented jointly with other TAs and the participating research and infrastructure communities. TA1 relates data quality to the definition of research data in the 4Memory RDM Strategy and to both metadata and data: the line between them is not drawn firmly but rather according to need, depending on disciplinary practices and data types.

If you do not want to miss any activities of TA Data Quality, sign in for our mailing list:
https://lists.nfdi.de/postorius/lists/ta1_data_quality_nfdi4memory.lists.nfdi.de/

Measures

Measure 1: Assessment of data-quality practicesPlease click button to expand or collapse content

The first step towards ensuring research data quality in historical research is assessing the status quo, allowing general guidelines and procedures for improving sustainable quality management to be developed. TA1 leads a working group to survey the needs and practices of – and available services for – data quality in our community. This effort includes different data types and all phases of the research data life cycle, considering existing guidelines and data-collection principles in RDM plans, methods, software, plausibility checks and data curation, provenance and ethical questions and data protection needs, data documentation, indexing with metadata, principles of standardised documentation, the use of metadata standards and normative vocabularies, the licensing and citation of data, and standards of scholarly practice. Together with TA4-M2, M1 also addresses the challenge of incorporating historical source criticism including hermeneutical and epistemological dimensions into data-quality measures. Apart from best-practices, obstacles to compliance with quality measures must be understood. Based on continuously collected problem stories, a systematisation of the surveyed requirements will further our understanding of what improvements are needed in data-quality practices. A tangible outcome of M1 will be a directory of existing technical recommendations, guidelines and standards of data processing for historical research.

Measure 2: Analysis of data-quality requirementsPlease click button to expand or collapse content

In M2, the data-quality status quo revealed in M1 enables analysing data requirements and evaluating actions for improving and ensuring their quality. Within this context, M2 also aims to make ethical issues an integral part of data-quality analysis via the FAIR and CARE-principles. This aspect will be complemented by NFDI4Objects in their Task Area 5 (Storage, Access and Dissemination) which will allow us to take into account both historical and archaeological research perspectives. A planned working group includes co-applicants, participants and other national and international partners (e.g. ICOM CIDOC) as intermediaries and experts for specific data-quality challenges. Data-quality requirements are analysed across the broad range of data types used in our community, based on problems named by our community and solutions developed in recent and ongoing projects within it. Also, data use and quality management issues in various contexts (editions, archives, collections, libraries, museums, memorial sites, etc.) are addressed. A catalogue of FAIR-compliant practices is systematically developed, and recommendations regarding the optimisation of workflows throughout the data life cycles are made. Particular attention is given to data transparency and traceability (e.g. data citation, licensing and persistence).

Measure 3: Development of a data-quality measurement systemPlease click button to expand or collapse content

M3 develops a measurement system for historical research data quality. An expert group representing various subcommunities assesses existing generic data-quality certification systems (whether project-based or institutionalised). An overview of existing quality schemes is systematised and synchronised through dialogue with the community. On this basis, a measurement system distinguishing various quality levels is developed, summarising the data-quality evaluation criteria and facilitating their application. The quality levels will enable the self-assessment of research data and facilitate a dialogue among data providers, data collecting institutions and data reusers. Existing peer-review standards and data-quality processes are evaluated in cooperation with TA5. Improving quality levels will enable interoperability (TA2 and TA3) and may qualify data for peer-review and scholarly publication (see TA5-M4).

Measure 4: Development of recommendations and guidelines Please click button to expand or collapse content

In close collaboration with international boards and committees, an expert group translates the findings of M1-M3 into recommendations, catalogues and guidelines for the creation, assessment and future certification of research data. The aim is to establish a permanent collaboration with existing international quality-assurance initiatives (e.g. GoFair, RDA). To that end, the expert group establishes a board for developing recommendations and identifies exemplary projects and work routines in which data quality is crucial for research and steps for developing suitable infrastructures for historical research. With regard to systematically implementing new project-management practices to improve workflows, this measure may serve as orientation for the DFG Review Boards and provide subject-specific information for the implementation of the DFG “Code” (Deutsche Forschungsgemeinschaft 2019), other third-party funding agencies and international (European) initiatives, committees and boards like CERI and RDA. Furthermore, the data-quality requirements for the 4Memory Data Space are defined. Tasks in this measure are carried out in collaboration with TA2, TA3, TA4 and TA5.

Measure 5: Development of editorial frameworks for data creationPlease click button to expand or collapse content

M5 aims to significantly lower the barrier of time-consuming data preparation through efficient workflows and a concept for dividing labour through editorial processes within an (organisational) framework applicable to existing institutional settings. Institutions in the consortium that engage in both RDM and traditional publication processes will jointly develop, implement and iterate editorial workflows for publishing research data. These data-editing processes – e.g. formal quality checks, format transformations, explicit documentation or metadata enrichment prior to publication – are developed with relevant actors in the key fields of research support, data management and publishing. Targeting formal data quality at an early production stage, the editorial procedures prime the peer-review process, focused on methodology and content, to be established by TA5. A curated network brings together actors from RDM and publishing departments. Sharing existing workflows and jointly identifying needs create prerequisites for a data editorial framework designed in a working group. The exploratory and preparatory work results in a service concept for data-editorial processes at the contributing institutions. M5 will allow the outcomes and concepts of M1-M4 to be reviewed in practice and, if necessary, adapted or extended. The collaboration of established editorial offices and newly installed RDM positions initiates institutional change.

Measure 6: Competence network: implementation of data qualityPlease click button to expand or collapse content

Crucial to all TA1’s activities is a constant dialogue with the community and all other TAs for the implementation and dissemination of data-quality measures. The guidelines will therefore be designed to meet different target groups. Particular attention will be given to smaller institutions, academic committees and learned societies as actors that structurally lack the capacities and infrastructure needed to develop and implement digital data-quality standards and practices at their institutions. Therefore, this measure provides RDM guidance and activities for data-quality training in coordination with TA4 (“train the trainer”).

Task area

Measures

Related Problem Stories

Services

Measures

Related Problem Stories

Measures

Related Problem Stories