Conceptual Framework for the Development of Digital Humanities Based on Artificial Intelligence

This study aims to develop a conceptual framework for advancing Digital Humanities (DH) through Artificial Intelligence (AI) technologies. A systematic literature review was conducted using the protocol proposed by Kitchenham et al., combined with thematic analysis and guided by the PRISMA framework. Articles were retrieved from the Scopus database (2015-2025). The initial search yielded 544 articles, which were reduced to 20 articles after applying inclusion and exclusion criteria and screening for direct relevance to AI-based digital humanities research. The findings indicate that research in AI-driven digital humanities is predominantly data-driven and computational, with quantitative methodologies accounting for approximately 60% of the analyzed studies. Machine learning models, particularly deep learning and transformer-based architectures such as BERT, T5, CNNs, and LSTMs, were widely applied to tasks including text and image classification, topic modeling, and automated cultural data analysis. These approaches were frequently integrated with Natural Language Processing (NLP) techniques and knowledge graph–based representations, enabling context-aware and semantically enriched analyses of large-scale, complex, and unstructured cultural and historical datasets. The results also suggest an emerging tendency toward multimodal and hybrid approaches. Based on the analysis, seven core components were identified as the foundation of the proposed conceptual framework: deep learning and neural network models; NLP and structural analysis; cultural and archival data analysis; challenges and limitations; opportunities and benefits; computational resources and data requirements; and community engagement and ethical considerations. Overall, the findings suggest that AI can effectively augment human-centered approaches, offering a structured foundation for advancing research in digital humanities.