COMPUTATIONAL SOCIAL SCIENCES
Digital Humanities
Analyzing cultural heritage through massive-scale AI models.
The Technical Measurement of Culture
Humanities meet High-Performance Computing. The digitalization of centuries-old archives requires infrastructures capable of not only storing petabytes of unstructured data but also semantically unlocking them through Computer Vision and Large Language Models (LLMs).
TEXT MINING & NLP
LLM-based Archive Analysis
Utilizing Large Language Models to analyze historical archives spanning centuries. We enable the identification of discursive shifts and semantic patterns across billions of text pages.
- Diachronic linguistic analysis
- Sentiment tracking across epochs
COMPUTER VISION
Pattern Recognition in Art
Leveraging image recognition algorithms to catalog and analyze artistic patterns in global museum databases. Automated style identification and provenance clustering on GPU clusters.
- Formal style classification
- Visual linkage between collections
Humanities Data Pipeline
| Research Focus | HPC / AI Method | Scientific Outcome |
|---|---|---|
| Historical Linguistics | BERT-based training on archaic text corpora. | Reconstruction of lost language stages |
| Art History | Deep Convolutional Neural Networks (CNNs). | Uncovering global artistic influences |
| Urban Studies | Social network modeling & Agent-based simulation. | Resilience planning for megacities |