Back to top

Introduction

Enhancing Historical Insights: Named Entity Recognition (NER) and Knowledge Graph Construction on the Chinese Student Weekly with ChatGPT
  • Introduction
  • Character Modification
  • Named-Entity Recognition (NER)
  • Knowledge Graph
  • The Chinese Student Weekly
  • GitHub Repo

    • Introduction

    The Chinese Student Weekly (henceforth The Weekly), a publication that thrived between 1952 and 1974 in Hong Kong, holds immense historical significance. Across its 1,128 issues, it served as a platform for established authors and student contributors, capturing the essence of its time and preserving invaluable cultural heritage.

    With a staggering few million characters spanning over two decades, understanding Hong Kong’s history through the Chinese Student Weekly poses a huge challenge. Our team is interested in how to summarize and visualize some interesting relationships within the Weekly. Our project seeks to address this by leveraging modern technologies, including pre-trained large language models (LLMs) and graph visualization libraries in Python. Specifically, we focus on Named Entity Recognition (NER) applied to the text in the years 1953 to 1962, complemented by a Knowledge Graph visualization.

    Should you have any suggestions and are interested in the dataset and the techniques, please contact us at data@cuhk.edu.hk.

    Members: CHIK Ching Ho (AIST/4), SHEK Tsz Him (AIST/4), SEN Yan Ho (AIST/4)