If you are having trouble with the sign in process, please <ContactLink>contact the SSH Open Marketplace team</ContactLink>.

If you are having trouble with the sign up process, please <ContactLink>contact the SSH Open Marketplace team</ContactLink>.

Subject

Social Sciences & Humanities Open Marketplace

Successfully submitted!

Any bibliography should be carefully defined  with regard to its excerption scope, processing methods and available data sources. 
The definition of the excerption scope is a key step and the following characteristics of documents should be taken into account: (a) their language, (b) country/place of publication, (c) year/period of publication, (d) type of document, (e) medium of document, (f) topic, and (g) in case of literary bibliographies primary / secondary content as well as any other relevant criteria.
Another important task is to decide on the methods for how to process the data. Most comprehensive, but time-consuming method is to process the data from the autopsy. In this case a record is prepared with a document in hand and the curator has full control over the content of the record. Another option is to re-use existing sources. This option requires IT resources to extract, enrich and harmonise large datasets (e.g. national bibliographies like German National Bibliography, digital libraries like Gallica, etc.). It enables processing of large datasets within a short period of time, but without the full control of the content of the record. Last option is the retroconversion, which means to convert analog sources (printed bibliography, card catalogue etc.) into the structured data. The most challenging aspect here is to extract as much information as possible either through manual retroconversion or an automated one (e.g. through HTR, OCR or NLP methods).
Based on these parameters, one can better define required technical capacities and skills to make informed decisions on the budget and sustainability of a bibliographic project.

Literary bibliographies have historically been units in academic institutions or libraries and have consisted of teams of catalogers (bibliographers) who manually (from autopsy) analysed documents and provided entries to publications or bibliographic databases. Nowadays, however, thanks to the advancements in automated data processing one could imagine establishing and running a literary bibliography without a cataloging team. Although having catalogers on board will probably always be desired.
Literary bibliographies in terms of their governance model might be considered projects supported by the work of departments, centers or teams in academia or libraries. In that case these units are responsible for processing a literary bibliography and it is recommended that such a task is a long-term commitment of an institution.
Good understanding of your governance model is needed from the early stages of the project, i.e. a model which suits your resources at any given time. A governance model is needed both to manage this kind of project, but also to successfully communicate your project’s needs and goals with your funders or management.
Regardless of a chosen legal or institutional status and governance model, a literary bibliography is a research infrastructure: a tool/device/resource which is a service supporting research (for academia, but also the society) and access to knowledge. Running a digital service demands good understanding of your user group and all aspects of digital service management. It is recommended for you be acquainted with IT service management standards which could be very helpful e.g. FitSM, ITIL, ISO etc.
Finally, it is the state-of-the-art approach to treat your bibliographic data as research data and it is helpful to start your project with a good understanding of Research Data Management, especially establishing and updating regularly a Data Management Plan could act as a checklist for you that will allow you to address necessary requirements.

The definition and implementation of the specific data model is of high importance for the further use of the data. The curator should take into account the needs of the specific community, cataloguing standards, and authority files as well as systems of persistent identifiers and descriptive thesauri and ontologies for the needs of the subject description.
The current bibliographic data processing is dominated by the standardised exchange coming from the librarianship sphere. The most popular format is undoubtedly MARC21 and its different variants. MARC21 is a specific hybrid format which might not be easy to understand outside the librarian community. Another option is BIBFRAME, LOD-based format coming from the librarian sphere as well. With regard to the interoperability, the common data standards can be used (e. g. Dublin Core). Specific option is to process the bibliography in the form of linked data/wikidata format. We would not recommend using proprietary data format unless there is no other option available, because it highly complicates the data interoperability.
Next to the format, the list of processed information (database fields) should be prepared. For most of the data, the relevant fields should be available in the standardised formats. There is a possibility to use nationally or locally defined fields as well. If any relevant field is not defined within the standard, it is possible to use e. g. note field in a consistent way.
For the needs of interconnection of the data it is crucial to implement the system of persistent identifiers. These identifiers are crucial for data interoperability and data enrichment from the external sources. National authorities files should be taken into account as PID general knowledge databases adapted for the given country. Individual authorities systems are mapped on the international Virtual Authority File (VIAF) which enables interconnection between various systems of national authority system as well as to wikidata. Subject description might be realised with the set of national authorities as well, but for the needs of specialised discipline bibliographies they might be too general and more granular hierarchy might be needed. The examples of such thesauri might be MESH or AGROVOC.
Finally, a crucial task is to understand the licensing of your data – both those that are being reused and those which are created manually. Proper licensing decision have decisive implications on future data reuse. For open digital literary bibliography Creative Commons licences could be recommended.

There are two main software options that support processing of a literary bibliography and they depend on whether a literary bibliography is being created fully automatically or the workflow includes manual creation of entries.
In the first scenario a software for discovery service is needed which does not need to have support for manual creation of entries (e.g. Vufind). It allows importing data into it and the user gets access to the data through a search engine, faceted browsing etc. These kinds of software solutions usually allow for machine access to data (e.g. OAI-PMH, Z39.50, API). For small scale projects solutions like Zotero or even spreadsheets might be of use.
The second scenario demands a software solution that supports manual creation of records and this could be an integrated solution. There is a number of Integrated Library Systems (ILS) on the market available under various business models or one can use open source solutions such as Koha or Folio.  More universal systems such as Wikibase create another option. Finally custom-made solutions are also prevalent, but building such a system from scratch is demanding and comes with sustainability challenges.

When your service is more defined and operating then more collaboration options open up. It is recommended for a literary bibliography to seek collaborations with research infrastructures on national (e.g. DARIAH ERIC nodes, OPERAS nodes), international (DARIAH, e.g. DARIAH Bibliographical Data WG, OPERAS, etc.) and disciplinary levels (literary researchers associations, other infrastructures e.g. CLS etc.). 
Services like literary bibliographies should be presented and discussed both on disciplinary events (conferences, workshops), but also on digital humanities and scholarly communication events. On many occasions poster sessions are a good way to present your tool. 
An operating service might also be an interesting resource for other parts of the research infrastructure ecosystem, e.g. metadata aggregators such as GoTriple or European Literary Bibliography. Making your data available through a protocol like OAI-PMH or API will allow other services to aggregate your data and reach wider audiences. Although, some service providers might also worry about the negative impact of such aggregation on the traffic on their platform. Reuse of your data by services can be, however, monitored if proper relationships with aggregators are established (e.g. allowing for access to some usage statistics).

Tomasz Umerle

Vojtěch Malínek

DARIAH-EU

This workflow describes the steps needed to establish a digital literary bibliography as an openly and freely available discovery platform. It is not a workflow for improving management of an existing service although it might be beneficial in such cases as well. Hence, this workflow defines steps in a chronological manner, although all of them are intertwined. The steps cover a series of key decisions and implementations starting from defining the scope of a bibliography, applying a governance model, decisions on data-related and software-related issues and establishing proper relationships with existing initiatives and projects. 
The workflow follows primarily the definitions, methodologies and experiences of long-standing projects of Czech and Polish Literary Bibliographies but is adaptable for the needs of any other bibliographic project.

How to create a literary bibliography?

Workflow steps(5)

1 Define scope, methods and sources

2 Define governance model and target user groups

3 Defining and implementing data models

4 Define and implement software solutions for data processing and access

5 Develop relationships with existing initiatives and ensure discoverability