[4O3-OS-29b] OS-29

[4O3-OS-29b-02] Dynamics of Creator’s Reputation

Time Series Analysis of Early Modern Japanese Literature using Japan Search

〇Kouhei Kasai¹, Ikki Ohmukai¹ (1. University of Tokyo)

Keywords:Time-series analysis, Digital archive, AI application, Data-driven humanities, Japanese literature

Epoch making authors' reputations are said to form over time, often be posthumous fame. While studies suggest that trends, anniversaries, and discoveries influence this process, the dynamisms remain unclear. However, with digital archives and modern analysis tools, we can identify the factors more easily. This study analyzed metadata from Japan Search, for 400 authors curated anthologies of early modern literature. Creating a time series based on birth years, we found metadata followed a power law distribution, few authors gained posthumous fame, while prolific writers showed reputation cycles decades. This pattern might indicate research accumulation and changing readership. Further research with larger and classified samples could identify additional triggers more accurately.