State-sponsored research and development (R&D), or “public research”, constitutes a complex nexus in which political mandates, administrative processes, and scientific innovation intersect. Capturing the dynamic operation of historical administrative systems poses significant methodological challenges. Digital Humanities (DH) methodologies, particularly Social Network Analysis within historical contexts, offer novel approaches for visualizing and interpreting such complex systems. Network-based studies in Digital Humanities have most frequently adopted person-centric approaches, focusing on individuals and analyzing their relationships, careers, works, or patterns of movement. While other document-centric approaches — such as citation networks in parliamentary texts or conceptual networks derived from topic modeling — have emerged, the analysis of administrative process flow through a document-centric lens remains relatively underexplored. This study proposes the construction and analysis of a document-centric network to visualize the circulation and structural logic of research administration itself. This approach is applied to the military R&D policy of the First French Empire, focusing specifically on the establishment and activities of the Commission des fusées incendiaires.

Established in 1810 by Napoleon I, the Commission des fusées incendiaires emerged as a strategic response to British naval dominance and Congreve rocket experiments. The commission embodied the Revolutionary legacy of civil-military cooperation by integrating military engineers and officers with scientists like Gaspard Monge. It was tasked with the systematic development and standardization of incendiary weapons. Although the commission produced limited immediate outcomes before the fall of the First French Empire, it provides an illustrative example through which to analyze the aspects of long-term artillery research policy in 19th-century France.

The dataset for this study is built by extracting structured metadata—including dates, actors, and technical subjects—from primary sources dispersed across various archives. The core corpus consists of relevant decrees, commission reports, and official correspondence between Napoleon I and commission members. These internal documents are cross-referenced with additional external materials—which will be further expanded in the future—such as contemporary newspaper reports on British Congreve rocket experiments and scientific treatises published in Britain, to reconstruct the historical R&D process.

This study utilizes a document-centric network model where archival units function as primary nodes. Rather than assuming a rigid linear progression, the methodology employs a typology of administrative triggers to categorize the functional links between records. For instance, a causation link is identified when reports on British technology prompt the emperor to issue a letter mandating the establishment of the commission. The process through which wartime exigencies or executive mandates lead to the commission’s concrete research plans is categorized as an implementation link, while a feedback link traces instances where internal reports are subsequently reflected in policy. By applying this framework, the study organizes disparate archival artifacts into an interconnected analytical model, providing a structured means to examine the procedural development of military technology within a bureaucratic context.

This study aims to achieve two primary goals. First, it seeks to establish a scalable foundation for classifying and visualizing the history of long-term military R&D projects. Given that state-sponsored military research, particularly in the artillery sector, is typically large-scale and spans extended periods, the objective is to implement a digital methodology capable of continuously connecting and expanding data to grasp its comprehensive history. Even when a specific commission is temporary, the underlying research activity often persists across various administrative contexts; this study focuses on providing a framework to integrate these disparate records into a continuous narrative. Second, the research intends to provide an efficient guideline for navigating complex and interconnected military technical records, which are often difficult to retrieve in standard archival environments. By clarifying relationships through the network model and visualization, this study aims to improve the efficiency of document search and retrieval. Although currently in its early stages and focusing on data from a single commission, our ultimate goal is to build a systematic dataset by tracing the administrative and scientific records of French artillery R&D policy throughout the first half of the 19th century.