Recap: DH Virtual Discussion Group – Fall 2022 Edition

In December we closed off our fifth edition of the DH Virtual Discussion Group, during which we heard from three different early career researchers in our community. The Discussion Group series was jointly organized this semester by Prof. Margherita Fantoli (KU Leuven Faculty of Arts), Dr. Sven Lieber (KBR), Prof. Julie M. Birkholz (KBR Digital Research Lab and Ghent CDH), and myself, Dr. Leah Budke (KU Leuven Libraries Artes). This semester’s edition was yet another engaging and inspiring series of talks and discussions centering on many different aspects of DH research.

In October, Paavo Van Der Eecken from the University of Antwerp introduced us to his PhD project and detailed the decisions that went into to developing the annotation methods for a large corpus of images from nineteenth century Children’s literature. For a recap of this session and to learn more about Paavo’s work, you can read the full recap blog post here.

November brought us an engaging presentation from Houda Lamqaddam on the concept of digital satellites. This was our largest group for the fall edition with an attendance of 27, which led to a dynamic Q&A moment at the end of the meeting. Houda’s presentation took us behind the scenes of her work on Project Cornelia. This project is a hybrid research engine which focuses on bridging art history and computer science. As such, it uses and develops datasets and develops data retrieval and visualization tools to use with seventeenth-century Flemish tapestry and paintings.

From this starting point, Houda’s presentation raised some important questions about the creation and long-term preservation of digital tools. As Houda explained, questions of maintenance and sustainability are at the heart of digital humanities research, especially when it comes to the development of new digital research tools.

To conceptualize the relationship between digital research tools and their role in specific research projects and beyond, Houda introduced the concept of “digital satellites” to us. There are three defining characteristics to this concept: digital satellites are (1) “built to support a specific goal including communication, exploration or analysis, (2) they are artificial in nature, i.e. they exist in highly non-digital spaces, and (3) they orbit around more grounded research material.”[1] Houda also emphasized the importance of the end of the satellite’s life, likening this phase to a type of “space debris.” This phase also deserves thought and preparation; according to Houda, we should be asking ourselves how digital research tools can be useful for us during this phase as well. Some of the steps that Houda defines to support long-term function of satellites include (1) archiving the design, (2) archiving process and requirements, (3) making code available in open access, and (4) considering removing tools that no longer function.[2]

Houda’s presentation sparked much thought and interesting discussion. While we often see how new tools are developed or how existing tools are integrated into various research projects during the DH Virtual Discussion Group presentations, Houda’s presentation encouraged us to think more critically about the long lifecycle of these tools. This is of particular importance when we think of digital scholarship as a concept that also encompasses research data management and scholarly publishing. Documentation and extensive archiving of research output, in this case in the form of digital research tools, is of key importance. Moreover, publishing code in open access has the benefit of not only allowing more access but also creating transparency and giving insight into the research process and development of specific tools.

During our final session in December, we heard from Laura Soffiantini about her research on the extraction of formulae from ancient Latin funerary inscriptions. As Laura explained during her presentation, the funerary inscriptions are typically written or carved into stone pottery or metal. This is a vast amount of material, but Laura’s corpus for this project was narrowed to texts written on tombstones. The pieces that Laura is particularly interested in in these texts—the formulae—are multi-word sentence(s) in semi-fixed or fully-fixed form(s) which are used as stock expressions. Links exist between these words and removing a word in the phrase cannot be done without altering the meaning of the phrase.

Laura was particularly interested in applying computational methods to this large corpus of textual data to test whether or not these methods hold any potential for the recognition and extraction of formulae. She relied on two approaches to test this: (1) a rule-based approach with a manually annotated dataset of funerary formulae and (2) a frequency-based analysis and extraction determined by word combinations.

Laura’s research gave us some insight into the application of different computational methods with a specific linguistic focus. The comparison showed that both methods hold potential for future application, but that neither was fail proof 100% of the time. As a result, Laura’s analysis is a great example of the importance of methods-based research. In addition to presenting us interesting research, it also helped our group to think about the digital methods we apply in humanities research and the way that we justify these decisions in our work.

On behalf of the organizers, I would like to thank all of our presenters from the fall 2022 edition as well as all of those who attended the sessions. We enjoyed learning more about these projects, catching up with and learning from others in our community, and welcoming new people into our DH circle! If you would like to join us for the spring 2023 edition of the Discussion Group, you can join our mailing list here. When we have the spring program finalized, we will also publish the details here on Scholarly Tales.

[1] Houda Lamqaddam, “In Search of Meaning: Thinking Information Visualization within Art History Research” (2022), 93.

[2] Lamqaddam, 98–99.


Print Friendly, PDF & Email