Jian is passionate about computational social science (CSS) and digital humanities. His research interest lies in applying big data computational, algorithmic methods, and AI applications to survey historical human development, cultural evolution, and linguistics. He is a data journalist and data scientist focused on Natural Language Processing (NLP) in English and Chinese language models. He speaks English, Mandarin Chinese, German, and studied 4 years of Latin.
Currently, he assists Dr. Ricardo Fernandes using complex systems theory to better understand patterns of historical causation in the rise and collapse of civilizations. His projects include researching the Roman Empire in collaboration with IBM Debater NLP Project and maintaining the IsoMemo Pandora project. His methodologies of interest are:
- Big Historical Data. Maintaining and documenting Pandora & IsoMemo database networks and web applications. These follow a non-hierarchical distributive model under FAIR and CARE principles. In addition, the goal is to apply artificial intelligence methods employed to retrieve data from unstructured sources.
- Modeling and Machine Learning. Developing novel open-source modeling tools to analyze historical data and made available via the Pandora & IsoMemo software platform. In particular, Bayesian modeling is employed to build complex hierarchical models able to integrate diverse sources of information. Data-driven and hypothesis-driven approaches are both employed to investigate historical causation. This includes the use of a variety of machine learning methods.
Software Applications & Digital Products