Here we share our model to predict whether the title in the table of contents in Ming literary collections is a letter title or not. For example, the model can predict with almost perfect accuracy that 上丞相康思公書 is a letter title but 劉先生墓誌銘 is not.

In collaboration with the Center for Chinese Studies 漢學研究中心 and our research assistant Katherine Enright, the China Biographical Database project created 438,000 records of training data for letter titles. We fine-tuned a transformer model with these training data. Queenie Luo queenieluo[at]g.harvard.edu wrote the code. For further information, please see THIS LINK.