At DTCH, method is approached as a layered process of scholarly production that makes research materials readable, modellable, analysable, documentable, and shareable in digital environments. The framework below outlines the core components of this process, from the relationship with sources to open production and shared learning.
We approach texts, manuscripts, archival documents, dictionaries and cultural heritage materials together with their source information, context and evidential value. The digital edition approach connects research material to a systematic chain of evidence through transcription, TEI-based encoding, variant management, the documentation of editorial decisions, metadata and versioning. In this way, sources are structured as documentable, traceable, comparable and reusable research objects.
We model the vocabulary found in historical and contemporary Turkish language sources together with terms, concepts, variants, equivalences, usage contexts, and source traces. Through term databases, ontologies, and knowledge graphs, we aim to develop reusable term–concept resources that make context visible for Turcological research and support comparative analysis.
We treat research materials that fall within the scope of Turcology as research components that can be organised, connected and reused through data models, metadata schemas, standards, interoperability principles and open documentation. We use context engineering as a methodological approach that makes the relationships between sources, evidence traces, data layers, editorial decisions and research questions explicit, traceable and manageable.
We use AI-based and computational methods in Turcology research in a critical, explainable and source-aware manner, including machine learning, large language models, natural language processing and agent-based workflows.
In computational analysis tasks such as term extraction, variant detection, named-entity recognition, morphological analysis, classification, comparison and data enrichment, the selection of models, tools and workflows is guided by the research question, data structure, source context and human supervision. Source-based information retrieval models such as RAG, human-supervised research automation and digital tool/prototype development processes are also considered within the same methodological framework.
At DTCH, research outputs are approached as documented production processes as well as final products. Open documentation, source visibility, versioned production, reuse, and sustainable publication are the core components of this approach. Through workshops, seminars, guides, sample datasets, and open learning materials, we aim to create accessible methodological spaces for researchers.