Conducted with an interdisciplinary approach within the scope of the TUBITAK 3005 Social and Innovative Solutions Research Projects Support Program in the Social and Human Sciences, the project also involves Prof. Dr. Sezen Karabulut from the Department of History at Pamukkale University Faculty of Humanities and Social Sciences, Dr. Merve Özdeş Demir, Research Assistant at the Department of Computer Engineering, Faculty of Engineering at Pamukkale University, and Dr. Mustafa İlter, Lecturer at İzmir Institute of Technology, as researchers.

Regarding the projects supported under the TUBITAK 3005 – Social and Innovative Solutions Research Projects Support Program for a period of 24 months, Project Coordinator Prof. Dr. Umut Karabulut stated:
"The project is an initiative to systematically reveal the everyday demands, expectations, and criticisms of the public in Türkiye during the Atatürk Era by analyzing a large number of reader letters published between 1923 and 1938 using AI-based natural language processing (NLP) and large language model (LLM) methods, combined with a bottom-up historical approach. It aims to understand how ordinary people’s political, economic, and social agendas influenced power-centered processes. The literature on reader letters has mostly emerged through historical and sociological studies using limited datasets, and a comprehensive, data-driven analysis of hundreds of letters has not been conducted. Furthermore, the use of NLP models adapted to the linguistic features of the Turkish language of the period is very limited, creating a significant gap in the literature. This project fills this gap and proposes a new conceptual and methodological framework both for social and human sciences and AI research."

Prof. Dr. Karabulut further explained that, using reader letters as the dataset and a bottom-up historical approach, the project will focus on the agendas of ordinary people during the Atatürk Era in Türkiye, and political, economic, and social processes will be explained from a bottom-up perspective:
"In this period, when modernization accelerated, nation-state formation took place, and state authority was strongly felt, to understand the agendas of ordinary people and their patterns in relation to government policies, the project will also utilize the contents of government agenda news published in relevant periodicals in addition to reader letters. These will be analyzed with natural language processing and large language models and subjected to a two-way examination: first, the effects of decisions made by the power mechanism on ordinary people will be followed; second, whether the demands, criticisms, and expectations of individuals influenced political, economic, and social policies will be observed. In this way, the project aims to achieve a holistic historical perspective. To capture the linguistic features of reader letters from 1923–1938 (words and expressions unique to the period), existing language models will be adapted using continual learning. This method will enable general-purpose language models to be sensitive to the unique structure of historical texts, increasing the accuracy and reliability of analyses while providing a historical adaptation framework for Turkish language models. To amplify the voices of broad segments of society, the dataset includes periodicals with diverse political perspectives. The dataset comprises Hâkimiyet-i Milliye (Ulus) and Cumhuriyet newspapers, which were close to the government and published throughout nearly the entire research period, as well as newspapers addressing different societal segments and published for limited periods, such as İkdam, Vakit (Kurun), Akşam, and Tanin, along with issues of Sebîlürreşâd and İctihad journals."