Publications
Gain greater insights into our diverse range of publications by delving deeper into their content and features.
Unveiling the Critical Nexus of Data Preprocessing and Transparent Documentation for Result Quality and Reproducibility in Digital History
Clodomir Santana, Michał Bojanowski, Demival Vasques Filho, Agata Błoch
(in release)
-
This study underscores the importance of adequate data preprocessing, transparency, and documentation in digital history research, showcasing how these often overlooked practices impact research quality and reproducibility. We present a topic modelling case study involving over 168,000 records of official correspondence of the Atlantic Portuguese Empire from 1640 to 1822 to illustrate how these practices, associated with standardised formats and metadata conventions, facilitate the sharing and reproduction of experiments. First, we evaluate the impact of data cleaning and preprocessing on model performance. Second, concerning model selection, we compare the performance of latent Dirichlet allocation (LDA), latent semantic indexing (LSI), and Gibbs sampling algorithm for a Dirichlet mixture model (GSDMM). We not only emphasise the often underestimated importance of data preprocessing and transparent documentation to strengthen the robustness of research and contribute to a culture of reproducibility, but also demonstrate the potential of topic modelling in digital historical studies, especially in the context of the Portuguese Atlantic empire. The study presents promising outcomes with carefully executed preprocessing, model selection, and parameterisation. Our findings encourage further exploration and refinement of the methodology employed.
Os Jesuítas e a Era do Algoritmo: Uma introdução à Análise de Sentimentos na Correspondência Colonial Ultramarina
Agata Błoch, Marcia Amantino, Clodomir Santana
(in release)
-
This article explores the intersection between digital technologies and historical research through a case study on the Society of Jesus in the Portuguese colonial empire from 1642 to 1822. Using tools from the field of Computational History, in particular sentiment analysis, the study analyses over 1,300 documents from the Portuguese Overseas Archive (AHU) on the Jesuits. The research applies an unsupervised ensemble approach to sentiment classification, dividing the corpus into bottom-up, top-down and horizontal correspondence categories. The results show a significant increase in negative sentiment during the expulsion of the Jesuits in 1759, reflecting the historiographical consensus on their persecution during the Pombalan reforms. The article emphasises the role of the historian in interpreting algorithmic results and underlines the need for critical engagement with both the data and the tools. It concludes with a plea for a collaborative approach between humanistic expertise and artificial intelligence to improve historical research.
Cracking the Historical Code
Agata Błoch, Michał Bojanowski, Clodomir Santana, Demival Vasques Filho
-
The chapter addresses a methodological approach to unstructured data and discusses the potential that structured data offers in the field of historical research. The dataset, which initially consists of textual content sourced from digital collections at the Portuguese Overseas Archives in Lisbon, undergoes a preprocessing phase that forms the basis for the extraction of structured data. The authors combine history, social sciences, and computer science to convert the correspondence repository into a machine‑processable form. This transformation is supported by an interdisciplinary strategy in which they weave together elements of effective content management, topic modelling, and social network analysis.
-
How to cite: Bloch, Agata, Michał Bojanowski, Clodomir Santana, Demival Vasques Filho. ‘Cracking the Historical Code From Unstructured Correspondence Corpora to Computational Analysi’. Models of Data Extraction and Architecture in Relational Databases of Early Modern Private Political Archives, May 2025. Crossref, https://doi.org/10.30687/978-88-6969-919-1/006.
Memórias Digitais
Demival Vasques Filho, Agata Błoch, Juciene Ricarte Cardoso, João Carlos Nara Júnior
-
Book is divided into four sections — O Digitalizado, O Digital, O Computacional, and O Historiador na Era do Big Data — each section explores the intersection of history and digital technologies, examining how digital tools are transforming historical research and memory preservation.
-
Section 1: O Digitalizado
-
Focuses on the democratization of historical information through digital archives.
-
Discusses digital collections like records of enslaved families in 18th-century Paraíba and the Projeto Resgate Barão do Rio Branco catalog, underscoring digital archives' role in preserving and connecting historical memories.
-
Highlights digital tools that enhance access to information on Brazilian independence, facilitating enriched historical text analysis.
-
-
Section 2: O Digital
-
Explores digital innovations in preserving historical memory.
-
Includes case studies from the National Archives' work in digital humanities, the Digital Memorial of Victims of Slavery in Pernambuco, and the Digital Memory of Jewish Refugees in Brazil.
-
Emphasizes the role of databases and digital platforms in providing public access to diverse historical memories.
-
-
Section 3: O Computacional
-
Examines computational perspectives in contemporary historical research.
-
Discusses the interface between users and historical knowledge production, addressing the biases in digital collections like the Hemeroteca Digital Brasileira.
-
Features quantitative and relational analyses of texts and social networks, such as the study of CAPES access requests and social networks among landowning families in Portuguese America.
-
-
Section 4: O Historiador na Era do Big Data
-
Investigates the role of historians amid massive digital data availability.
-
Explores how historical practice is evolving with digitalized sources and large datasets, allowing for new perspectives and interpretations by connecting diverse data.
-
Argues for the critical value of digitalization in broadening access to colonial archives, enabling more comprehensive and diverse historical insights.
-
"Od historii cyfrowej do obliczeniowej. Studium przypadku cyfrowej historii kolonialnego imperium portugalskiego."
Agata Błoch and Clodomir Santana
-
The chapter explores various approaches to digital history with an emphasis on computational methods.
-
The goal is to show that computational approaches are accessible to researchers across different historical periods and sources.
-
The central question addressed is whether the research aims to obtain digital results from data processing or to experiment with algorithms.
-
Each research path—obtaining digital results or experimenting with algorithms—has its own justification in the context of digital history.
-
The first approach (obtaining digital results) can be pursued independently by historians.
-
The second approach (experimenting with algorithms) typically requires support from an interdisciplinary team.
-
The chapter includes a case study on the colonial Portuguese Empire to demonstrate computational history in practice.
-
Two models are showcased: sentiment analysis and topic modeling.
-
The chapter presents the functionality of these models, including graphs, codes, and practical applications for both digital research and historians.
Cite as: Agata Błoch and Clodomir Santana. "Od historii cyfrowej do obliczeniowej. Studium przypadku cyfrowej historii kolonialnego imperium portugalskiego." In:
Człowiek twórcą historii. Warsztat nowoczesnego humanisty historyka na progu XXI, eds. Cezary Kuklo i Wojciech Walczak, v. 5., Uniwersytet w Białymstoku, 2024, 23-50.
"Introduction to historical (social) network analysis"
Demival Vasques Filho
-
Discussed landmark papers in the evolution of network research.
-
Historical network analysis is a discipline in which network analysis is used for studying history, but there are no new methods or concepts that distinguish it from traditional social network analysis.
-
Leonhard Euler's theorem in 1736 laid the foundations of graph theory in mathematics by identifying two conditions for a walk across a graph.
-
In 1934, Jacob Moreno and Helen Jennings used nodes and edges to create a sociogram to represent social ties, which is considered the birth of social network analysis.
-
Paul Erdős, Alfréd Rényi, and Edgar Gilbert made essential developments on random graphs in the 1950s that gave the basis for researchers to explore phase transitions and expected properties in the structure of these random graphs.
-
In the 1970s and 1980s, several seminal papers on social networks were published, including Mark Granovetter's "The strength of weak ties," which drew attention to the fact that weak ties, like acquaintanceships, can be better for accessing information than strong ties.
-
Weak ties facilitate the diffusion of things on social networks, connecting communities and allowing these things to spread further.
Networks from archives: Reconstructing networks of official correspondence in the early modern Portuguese empire
Agata Błoch, Demival Vasques Filho, Michał Bojanowski
-
The study focuses on a large corpus of administrative correspondence from the Portuguese Empire, spanning over 220 years and comprising almost 170,000 documents.
-
Traditional archival work can be time-consuming and limit the scope of network studies. To overcome this, the study proposes a modern natural language processing approach for extracting network data from unstructured textual information.
-
The resulting dataset includes additional information such as the occupation, administrative affiliation, and geographical location of senders and recipients, making it a valuable resource for historians and social network researchers.
-
Preliminary network analysis of the dataset reveals its potential for investigating the political and social evolution of the Portuguese Empire during the reign of seven Portuguese monarchs.
-
The study's proposed approach is not only applicable to this dataset but also adaptable to other contexts, languages, and types of historical archives, making it a valuable contribution to the field of historical network studies.
The “Miserable Vassals” of the Empire: The Androgynous Codes of Behaviour of Black and Indigenous Peoples in Late Colonial Brazil (1775–1808)
Agata Błoch
-
Combined linguistic coding and androgyny category to analyze discursive patterns in negotiations between Indigenous and Black individuals and colonial authority in Brazil (1755-1808).
-
Examined language of identity and codes of behavior to understand how Black and Indigenous subjects perceived freedom, social condition, slavery, and colonial power.
-
Used intersection of gender, social status, and race/ethnic group to measure androgyny.
-
Examined culturally expected feminine and masculine attributes reflected in individuals' self-description.
-
Discussed "miserable vassals" and functional aspects of androgynous codes of behavior.
Slaves, Freedmen, Mulattos, Pardos, and Indigenous Peoples. The Early Modern Social Networks of the Population of Color in the Atlantic Portuguese Empire
Agata Błoch, Demival Vasques Filho, Michał Bojanowski
To create and visualize social networks of people of color and understand their position in a colonial structure, there are three possible approaches:
quantitative research focusing on specific social actors, analyzing colonial official correspondence mentioning people of color, and qualitative research examining personal narratives and relationships mentioned in petitions sent to the colonial administration. Each approach provides valuable insights into the historical context and position of people of color in the colonial system.
“The Free and the Enslaved. Voices of the Subalterns in the History of Portuguese Empire”
Agata Błoch
-
Network perspective of network is a way of looking at human relationships, aimed at building a network of contacts.
-
Individuals with similar characteristics occupy the same position in the network structure, but their influence on its co-creation may differ.
-
Applying this perspective to the analysis of the Portuguese colonial society, it can be viewed as a network society, where the elements define each other.
-
The structure of the network impacted the behavior of subaltern groups, cultural, social, and identity changes, leading to a sense of attachment to Portuguese colonial structures.
-
The Portuguese administrative correspondence and selected petitions of African and indigenous people were analyzed using an egodocumental approach, grouped thematically, not by ethnic, racial, or geographical criteria.
-
The social networks of these groups reflected their awareness of their existence and the need to build diverse relationships.
-
Freedom can only be obtained through interactions with others.
-
The agency of the subalterns dependent on the contact they could establish and maintain with the white colonizers, who could be their masters, soldiers, clergymen, colonial officials, and others belonging to the elite.
-
Indigenous and African slaves had similar awareness of their surroundings and the people forming the network, regardless of whether they were newly arrived in Brazil or their descendants.
-
Defining their position through the network was crucial to these groups.