Libération des données

An Open Science Approach to the Curation and Analysis of the XVIIIe siècle: bibliographie

Christof Schöch
Trier Center for Digital Humanities, Germany

Digital Humanities Conference 2024
Washington, DC, USA – George Mason University

08 Aug 2024

Introduction

Overview

• The XVIIIe siècle: bibliographie
• An Open Science approach
• Data conversion
• Descriptive analysis
• Patterns of collaboration
• Limitations and next steps

• In the following, I would like to answer a few questions:
• What is this bibliography that I’ve worked on?
• Why and how use an “open science approach” to curating it?
• What kind of data conversion was required?
• What’s actually included in the bibliography?
• More specifically, what patterns of co-authorship and co-editorship can be discerned?
• What are some of the issues with the dataset, or the analysis so far?
• That’s plenty, so let’s dive right in.

The XVIIIe siècle: bibliographie

What is the XVIIIe siècle: bibliographie?

• Published by Canadian scholar Benoît Melançon
• Covers research on the Eighteenth Century
• Appears in monthly instalments since 1992 (!)
• Now includes 72.000 references
• All data for 1992-2022 was published in bulk in early 2023

See Melançon (2023) and Melançon (1992-2024).

Compared to MLA International Bibliography

	BIB18	MLA
items, 1992-2022	64.400	102.000
items, 1900-2024	72.200	180.000
items in French	74%	~10%
items in English	21%	~70%

• Reference bibliography: MLA
• MLA is 2x to 3x larger
• But: Language coverage is complementary:
• MLA has 18k in French
• BIB18 has 53k in French

An open science approach

Why Open Science

• Open Science is science done right
• Making data freely available, but not in standard formats, is going only half the way
• The goals don’t justify the means 🤷

Image credit: Melanie Imming, Jon Tennant. See: Tennant (2019).

How is this Open Science?

• Open data: Data, code, visuals in Github repository
• Open source: Only freely-available software
• Open process: Only algorithmic approaches – no Gephi 😢
• Open access: Published using Jupyter Notebooks in Quarto

See: christofs.github.io/BIB18/.

Data conversion

Data formats involved

• Benoît Melançon
  • Entry of data via web form
  • Managed in a FilaMaker DB
  • Monthly exports to HTML for the web display
  • 2023: Bulk export to CSV
• My work
  • CSV to BibTeX using Python for import into Zotero
  • Manual and semi-automatic corrections
    
  • Exports from Zotero: BibTeX, XML-RDF, BibJSON

Descriptive analysis

Number of entries per year

Site de Benoît Melançon

Prevalence of publication types

Most frequently-mentioned scholars

Most prevalent languages

Most prevalent publishers

Words frequently mentioned in titles

Patterns of collaboration

Co-authors: percentages and network

• French-language Eighteenth-Century studies is a single author field

Co-editors: percentages and network

• BUT! French-language Eighteenth-Century studies is a multiple-editor field!
• And, note also that edited volumes are immensely important in the field.

Limitations

Some limitations

• Data curation: self-selective curation: biases?
• Data model (Zotero):
  • Multiple chapters in edited volume not modeled as “part-of” relationship
  • Monograph vs. edited volume vs. scholarly edition
  • Gender of authors and editors not modeled explicitly
• Zotero as a tool: unwieldy for collection of 64k items
• Data quality: duplicates, languages, missing abstracts

Many thanks! 😺

References

Melançon, Benoît. 1992-2024. “XVIIIe siècle: bibliographie.” Site de Benoît Melançon, 1992-2024. https://benoitmelancon.quebec/xviii/biblio.tdm.html.

———. 2023. “Libération des données.” Site de Benoît Melançon. https://benoitmelancon.quebec/xviii/donnees_biblios_1_550.html.

Tennant, Jon. 2019. “Open Science: Just Science Done Right?” Figshare. https://figshare.com/articles/presentation/Open_Science_Just_science_done_right_/9759353.