Figure 1 with annotations

Today at 3:30pm Lisbon Time (busy day for me!) I will also be presenting our work, "A Study of Imagery in Franz Kafka's Novel The Trial Through Illustrated Editions" with my colleague at the University of Missouri, Professor Carsten Strathausen at #DH2025.

Presenting our work, “Text Mining Gender Depictions in Northern Wei (386–539 C.E.) Epitaph Verses,” with Professor Seiko Ochi (Meijo University, Japan) today at 2 PM Lisbon time at #DH2025! Please join us in person or online if you can!

Invited by Shiraz University's Digital Humanities Association, I'll be presenting my work with @tedunderwood.me on classifying English poetry tomorrow at 8:30am US Central Time via Google Meet. Excited to be part of the growing global conversation on DH, including students and scholars from Iran!

Presenting our work with @rongqianma.bsky.social and Heather Moulaison-Sandy at iConference #iconf25 tomorrow at 9 AM! We'll share our initial findings on how AI is discussed in digital humanities journals. If you're in Bloomington, we'd love for you to join us! Full paper: doi.org/10.47989/ir3...

New poster on my office door celebrating the USA's 250th birthday! As noted by its creator @ryancordell.org , the references to "tyrants" and "kings" clearly link to the 18th-century British monarchy—not to any self-proclaimed king of the 21st century.

Presenting our work, "Interplays Between Materiality and Content in Book History: Evidence from 16th–19th Century Chinese and English Books" (coauthored with Yuqi Chen, Ryan Dubnicek, @ryancordell.bsky.social & Stephen Downie), at #DH2024 tomorrow morning at 8:30 AM. Hope you can join us!

My dissertation acknowledgements.

The relationship between poetic prominence and semantic accuracy or prosodic accuracy does not demonstrate a consistent pattern. The strengths of semantic and prosodic characteristics are primarily associated with specific poetry categories and are not strongly influenced by author canonicity. (5/6)

The difference in accuracies between the two classifiers for each category shows a unimodal pattern, indicating no clear division between categories more driven by semantic features and those more driven by prosodic features, as we would expect a bimodal distribution if such division exists. (4/6)

When classifying each category against all other categories, sonnets, heroic couplets, concrete poems, haiku, and limerick are better distinguished when classified by the prosodic classifier, while all other categories display superior distinction when classified by the semantic classifier. (3/6)

We built machine learning models based on semantic features (lexicons) and prosodic features (meters and rhymes) independently. Although the classifiers exhibit similar overall performances, their performances on individual categories differ significantly (in ballads, heroic couplets, etc.) (2/6)

Looks like they are just using titles

Presenting “Structural Characteristics in Historical Networks Reveal Changes in Political Culture: An Example From Northern Song China (960–1127 C.E.)” at 11:30am Paris Time at #CHR2023 with Professor Song Chen, Yuqi Chen, and Professor Jana Diesner. Full paper at: ceur-ws.org/Vol-3558/pap.... 1/5