Donghan Bian

PhD Student

Team

Digital Methods in the Humanities and Social Sciences

Campus

Paris

Contact

donghan.bian@epita.fr

Research Area

Digital Humanities
Natural Language Processing
Text Analysis
Information Retrieval

Me in brief

I am a PhD student at the École nationale des chartes and in the Digital Methods in the Humanities and Social Sciences research team at EPITA, under the direction of Édouard Vasseur and Marie Puren.

My research lies at the intersection of natural language processing and digital humanities. I develop and evaluate computational methods for exploring large, heterogeneous collections of historical documents, with a particular interest in tracing events and public discourse across sources.

Short Bio

My academic background combines linguistics and digital humanities. I studied French language and literature at Zhejiang Gongshang University, completed a master's degree in linguistics at UCLouvain, and subsequently pursued a master's degree in digital humanities at the École nationale des chartes.

Research

Research Interests

My research focuses on the methodological challenges of analyzing historical and social phenomena through large-scale textual data. I am particularly interested in:

  • Historical event tracking: identifying references to the same events across documents that differ in vocabulary, genre, and historical context.
  • Natural language processing and information retrieval: developing robust, interpretable, and traceable approaches to searching noisy and heterogeneous archives.
  • Computational discourse analysis: studying public discourse, political communication, and the circulation of narratives through text.
  • Digital humanities: connecting computational methods with historical interpretation and source criticism.
  • Language model representations: investigating what internal representations capture and whether they offer useful signals beyond conventional lexical and semantic retrieval methods.

Publications

A list of my publications is available on: