Abstract of the track

The section “Digital Humanities and Modeling of Complex Subject Areas” is about computational and formal methods for representing and analyzing complex subject areas. Many fields deal with data that is highly interconnected and complex, and such data cannot be fully described by simple flat or tabular models. Neural networks and large language models (LLMs) are powerful tools for processing this kind of data, but they can produce hallucinations and are hard to interpret. This limits their reliability as stand-alone models of complex domains.

The section brings together researchers who combine neural approaches with formal structural models, such as complex graph models (hypergraphs, hypernetworks, metagraphs), to build more reliable and interpretable models of complex subject areas. Contributions are welcome in any complex subject area, such as digital humanities (for example, media discourse analysis), as well as natural sciences (for example, biological and medical applications), where neural networks, LLMs, and complex graph models are used to represent and analyze complex data.

Subject & Research Domains

Subject: Digital Humanities and Modeling of Complex Subject Areas

• Digital humanities, text and media discourse analysis
• Complex graph models and hybrid neuro-symbolic approaches
• Neural networks and large language models for complex domain modeling
• Biological, medical, cultural, and historical data

Topics

Topics include, but are not limited to:

• Modeling of complex subject areas using complex graph approaches (hypergraphs, hypernetworks, metagraphs)
• Application of neural networks and large language models (LLMs) for modeling complex subject areas
• Hybrid neuro-symbolic approaches combining LLMs with formal graph-based structures
• Limitations of neural methods (hallucinations, interpretability, verifiability) in complex domain modeling
• Formal and computational models of text and media discourse
• Knowledge graphs and ontologies for complex subject areas
• Computational and graph-based modeling in biology and medicine
• Digital methods and information systems for humanities, cultural, and historical data
• Multimodal data modeling (text, image, video, biomedical data) using neural networks and complex graphs

The topics listed above are not exhaustive; other works related to digital humanities and modeling of complex subject areas are also welcome.

Recommended Invited Speakers

Yuriy Gapanyuk
Bauman Moscow State Technical University

Irina Karabulatova
Heilongjiang University; Moscow State University

Sun Qiuhua
Heilongjiang University, Director of Institute of Russia and China

Olga Lebedeva
Great Peter Sankt-Peterburg State Polytechnical University