While Digital Humanities (DH) scholarship has extensively engaged with Western and Japanese graphic narratives, Chinese lianhuanhua—pocket-sized comic books—remain largely overlooked. This proposed short paper reports on a major work-in-progress: the digitization and metadata database creation for a unique collection of these ephemeral materials. First, we outline the conceptual framing and digitization strategies for a collection of over 3,000 lianhuanhua. We address the preservation challenges posed by their material instability and the critical need for an openly accessible digital infrastructure. Second, we detail the technical implementation of our project’s metadata backbone. We present a case study on digitizing a comprehensive bibliographical catalogue containing publication data for over 35,000 lianhuanhua titles. We outline our reproducible, semi-automated workflow, which utilizes PaddlePaddle OCR for complex Chinese character extraction, alongside a hybrid approach of Large Language Models and Python (pandas) to semantically parse and structure the disorganized text into a 38-column relational database. This rigorous standardization resolves significant hermeneutic challenges in data management. Ultimately, building this massive, machine-readable dataset paves the way for advanced DH applications—such as network analysis and geospatial mapping but also for other kinds of catalogue digitization—to answer foundational questions about the canon, mass-print popularity, and longue durée evolution of Chinese socialist comics. By bridging physical artifacts with digital information spaces, this paper offers a scalable workflow relevant to the broader DH and global cultural studies communities.

This short paper reflects on a computer vision experiment examining gender representation in Chinese comics (lianhuanhua), small illustrated storybooks widely circulated in twentieth-century China for entertainment, education, and political communication. While prior scholarship has focused on close readings of individual works, the medium has rarely been analyzed at scale, leaving broader patterns of visual and textual representation underexplored. To address this gap, the study combines structured annotation with computational modelling to investigate how female figures are visually constructed and narratively positioned across panels.

The central methodological decision involved model selection and experiment design. Rather than maximizing the number of detected figures, the project prioritizes interpretive rigor and contextual accuracy. Supervised object detection was avoided due to potential gender bias and the highly stylized nature of lianhuanhua imagery; instead, image clustering was applied to annotated character regions. A multimodal annotation framework cross-references visual features with narrative text, and large language models assist in identifying gendered cues, enhancing efficiency while preserving interpretive oversight.

The process generated a structured, reproducible multimodal dataset linking image regions, textual content, and annotations. Preliminary findings reveal both visible imbalances and subtler phenomena, such as textual mentions of characters absent in illustrations, highlighting gaps between narrative acknowledgment and visual presence. An ongoing phase explores AI-assisted attribution of objects to female figures.

The study contributes to Digital Humanities by proposing a bias-aware experimental workflow, producing a reusable multimodal dataset, and demonstrating how computational analysis of textual–visual alignment can generate new insights into gendered visibility in historical media.

Property regimes reshape cities through boundary making, subdivision, consolidation, and the allocation of development rights, yet the spatial traces of these operations, especially parcel geometries and their afterlives in built form, remain difficult to reconstruct and compare at scale. Evidence is dispersed across historical maps and plans, cadastral surveys, land registers, court and governmental files, and architectural traces, with discontinuities in scale, terminology, projection, and status. This paper presents a method-first spatial humanities workflow that treats property lines as spatial evidence and translates them into a comparable sequence linked to ownership indications and built outcomes.

Using Jaffa and Tel Aviv as a case study area, the method begins from about forty historical maps and plans spanning the nineteenth century to the present, assembled as high resolution scans and stabilized with identifiers and provenance records. Maps are georeferenced in GIS with documented ground control points, transformation parameters, and uncertainty notes. Parcel boundaries and subdivision lines are traced into vector layers using a minimal schema emphasizing provenance, date, feature type, and confidence. Traced parcels are then linked, where possible, to land registration and selected juridical and administrative documents, treating cartographic and legal materials as complementary registers of claim making. Parcel evolution is analyzed for morphological signatures, then read alongside planning approvals and building histories for selected sites.

Early results include a pilot boundary atlas that makes long term parcel change traceable and reveals lags between legal and physical change, enabling systematic connections between boundary operations and architectural form.

The article presents the results of a pilot project investigating the potential of automated manuscript analysis using large language models (LLMs) and HTR tools. The aim is to develop and evaluate a practical workflow, compare available solutions, and formulate good practices for future projects involving fifteenth-century handwritten sources. As a case study, a fifteenth-century formulary manuscript (BJ rkps 42), written in Gothic cursive by multiple scribes, has been selected. The manuscript poses significant challenges due to paleographical variability, numerous abbreviations, and linguistic complexity resulting from the interaction of Polish and Italian traditions.

The project consists of three modules. (1) Automated text acquisition (HTR): different infrastructures and engines are tested, along with various training strategies. The use of LLMs for post-editing and correcting transcription errors is also examined. (2) Translation and philological processing: several LLM are compared in terms of semantic accuracy, grammatical correctness, and sensitivity to administrative formulae, with particular attention to prompting strategies. (3) Information extraction: generation of regesta and extraction of structured data (persons, places, institutions), including TEI annotation and linking to external authority files.

The results make it possible to assess the real potential of automation and to propose a scalable workflow for processing fifteenth-century manuscripts.

Historical maps represent rich cultural and linguistic artefacts, yet their textual components remain underexplored in digitization initiatives. This study investigates the integration of Optical Character Recognition (OCR) and Natural Language Processing (NLP) methods into a broader digitization project at a university library, aimed at improving accessibility and scholarly engagement with a large cartographic collection. The project combined a user-friendly map portal with computational workflows to extract, transcribe, and analyze textual elements such as place names, legends, and paratextual data. A multi-stage workflow was developed, incorporating text segmentation, semi-automatic transcription, NLP-based categorization, and corpus-based analysis. Several open-source OCR tools including eScriptorium, PERO OCR, Tesseract, OCR4all, and OCRmyFiles were evaluated for recognition accuracy, handling of heterogeneous layouts and typefaces, and adaptability for manual correction. Challenges encountered include uneven text distribution, variable orientation, low contrast, overlapping graphical elements, and damaged sources, all of which constrained automated recognition. Segmentation quality proved a critical bottleneck, emphasizing the need for hybrid approaches that combine automation with targeted manual intervention. Corpus-based visualizations using Voyant Tools supported exploratory analysis, comparative evaluation of OCR output, and metadata enrichment, though limitations remain in representing multi-word geographical terms. The study also explores incremental refinement of domain-specific transcription models to improve recognition performance. Overall, the research highlights both the potential and limits of automated text processing in digital cartography, advocating a pragmatic hybrid approach that treats cartographic text recognition as a selective, research-driven process rather than a fully scalable solution.

Mass digitization and the expansion of born-digital archives have multiplied the availability of historical sources, yet effective reuse remains constrained by fragmentation across formats, metadata schemas, OCR conditions, and institutional infrastructures. Rather than a mere shortage of tools, this “Babel” condition reflects an infrastructural and epistemological fragmentation of historical workflows. Interoperability cannot be assumed as an automatic by-product of digitization; it must be designed as a governable, traceable workflow condition. This short paper conceptualizes AI not as an autonomous analytical engine, but as a human-governed mediation layer operating between heterogeneous source ecosystems, standards, and research practices. AI is positioned as an orchestration mechanism that assists transitions between workflow phases, particularly where friction arises between formats, schemas, and languages, while preserving source criticism, tool criticism, and expert validation. A modular six-step workflow is proposed: heterogeneous ingestion; OCR/HTR retrieval and correction; normalization; semantic enrichment; remodeling into interoperable structures (CSV/JSON/graph-ready outputs); and analytical production (networks, timelines, geodata). The focus is methodological rather than technical, emphasizing interoperability bottlenecks such as metadata normalization and schema mediation.

The argument is demonstrated through a comparative proof-of-concept on two historical serial corpora: French Red Cross bulletins hosted on Gallica and researcher-digitized Italian Red Cross bulletins. Evaluation adopts workflow-oriented criteria—time-to-integration, reduction of manual mediation, adaptability, traceability, and qualitative error taxonomy. The paper proposes AI-mediated interoperability as a reusable, inspectable, and methodologically accountable strategy for historically grounded Digital Humanities research.