Background. Taiwan Biographical Database (TBDB) is a database collecting biographical information on historical personages in Taiwan. It provides functions including relational database (RDB) and full-text search, social network analysis (SNA) and geographic information system (GIS). The primary purpose of TBDB lies in establishing a digital-humanities-oriented system that inspires historians to explore Taiwan history. Objectives. Through semi-automatic named entity recognition (NER) from the full text of biographies, TBDB assists historians to construct networks of social relationship. However, the full text of biographies may not describe all social relationship of historical personages. In consideration of the fact that historical photographs were usually taken in formal occasions, historical photographs may be utilized to uncover additional relationship. In view of this, this project proposes a face detection function in TBDB that utilizes the OpenCV Library to detect faces of historical persons in old photographs. Furthermore, it employs hierarchical agglomerative clustering (HAC) to combine fragmentary social networks. In addition, this project attempts to establish an ontology that can be used to represent the information of historical persons explicitly and useful for the reuse and exchange between different systems. Results. An experiment using 45 historical photographs shows that the face detection function could achieve average 98.16% recall, but with a low precision. In spite of the poor average precision, a convenient user interface has been implemented in TBDB that facilitates users to delete false-positive faces in old photographs. On the other hand, the average high recall allows users to retrieve nearly all faces in old photographs. Furthermore, with HAC to integrate social relationship found in biographies, social relationship detected from historical photographs, and even social relationship given by external sources, TBDB could produce comprehensive social networks to inspire historians. In addition, an instance that applies the proposed ontology is presented.
|Effective start/end date||2018/08/01 → 2020/12/31|
- Digital Humanities
- Taiwan Biographical Database (TBDB)
- Social Network Analysis (SNA)
- Face Detection
- Hierarchical Agglomerative Clustering (HAC)
Explore the research topics touched on by this project. These labels are generated based on the underlying awards/grants. Together they form a unique fingerprint.