Digitization and Metadata Generation Process
We digitized 94 out of 156 total issues of USSR in Construction from the Wende Museum's collection, accounting for roughly 60% of the magazine's complete publication history spanning 1930-1949. Our focus on 1930-1935 captures the critical early Soviet period encompassing the first and second five-year plans before the Great Purges intensified.
Digitization workflows included photographing, color correcting, and making the magazines digitally navigable using Issuu. Multiple technical approaches for visual and textual data extraction were employed post digitization. OCR converted scanned text into machine-readable format, enabling large-scale analysis across the corpus. We prioritized back pages and end matter to identify photographers, editors, and production credits—essential for mapping networks of cultural producers.
Data cleaning through OpenRefine standardized inconsistent naming conventions, dates, and geographic references. This revealed translation and transliteration challenges with the materials, demanding careful attention to authority records and consistent naming practices.
Subject Headings and Authority Records
We developed a hybrid approach combining Library of Congress Subject Headings (LCSH) with locally developed vocabularies for themes: industrial development (steel factories, infrastructure), social themes (collective farming, public health, women's roles), geographic locations across the USSR, and cultural figures (photographers, writers, designers).
Name authorities presented challenges; Editor-in-chief A. Khalatov appears as "A. Halatov" or "Halatoff" across issues; photographer M. Alpert sometimes appears as "M.V. Alpert." We consulted VIAF records to establish standardized forms. Geographic authorities required historical context, with locations like Leningrad (now St. Petersburg, formerly Petrograd) needing temporal notation.
Organizing the Project
Rather than focusing narrowly on specific visual elements, we adopted a broader thematic approach creating a comprehensive dataset for diverse research questions about modern propaganda campaigns. This required developing frameworks for categorizing propaganda methods—visual versus textual approaches with further subcategorization within each type.
The temporal focus on 1930-1935 captures the optimistic period of Soviet propaganda before political purges intensified. This timeframe encompasses the first five-year plan (1928-1932), during which USSR in Construction promoted state-led efforts, and the beginning of the second five-year plan (1933-1937), focused on heavy industry and technological modernization.
File management presented ongoing challenges with large file sizes complicating data transfer between systems. We maintained both individual image files for OCR processing and consolidated PDFs for broader analysis, balancing technical requirements with research accessibility.
Why a Network
Network visualization effectively reveals complex interconnections between themes, locations, people, and temporal developments characterizing Soviet propaganda. Unlike traditional archival access systems, network graphs enable exploration of non-obvious connections and influence patterns across the magazine's content.
The magazine presented a network of cultural producers; writers Maxim Gorky, Lev Kassil, and Isaak Babel; photographers Arkady Shaikhet, Max Alpert, and Yevgeny Khaldei; designers El Lissitzky, Alexander Rodchenko, and Solomon Telingater. Understanding collaborative patterns across issues and themes requires visualization tools representing multiple relationship types simultaneously.
Network Structure and Connections
The network structure treats themes, locations, and issues of the magazine as interconnected nodes, with edges representing temporal co-occurrence, geographic proximity, collaborative authorship, and thematic similarity. This enables complex research questions about how Soviet propagandists defined and organized information, how different USSR regions were represented across issues, and whether themes like collective farming or industrialization varied over time.
The network incorporates geographic analysis, mapping locations referenced or depicted in magazines and tracking infrastructure development across the USSR. This spatial dimension reveals how propaganda narratives were constructed around specific places and how the Soviet state projected power across its territory.
Technical Implementation and Presentation
We evaluated multiple visualization platforms—Cytoscape, Flourish, and Nodegoat—seeking to balance analytical sophistication with user accessibility. We found that Kumu provided robust network analysis, simple implementation, and metadata import capabilities. We also appreciated Kumu's capacity to embed images directly into network nodes.
We created collaged banner images from visual elements from the magazines throughout the site that echo the magazine's distinctive layouts. This maintains visual consistency with the source material. The network and accompanying Python visualizations allows users to filter and search by topic, explore temporal changes, and trace geographic focus across the collection, creating a rich environment for research of these materials. We have also made our complete dataset of cataloged USSR in Construction magazines accessible to the public.
View our dataframe here!