An Experiment on Dynamic Extraction of the Human Relationships on the Web Using Closeness Centrality

• Published in: 2022 International Conference on Computational Science and Computational Intelligence (CSCI) pp. 839 - 844
• Main Authors: Ichikawa, Ayaka, Onuma, Ryo, Nakayama, Hiroki, Kaminaga, Hiroaki, Miyadera, Youzou, Nakamura, Shoichi
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-12-2022
• Subjects: Closeness centrality; Computational intelligence; Dynamic extraction of human relationships; Feature extraction; Human-relationship network; Observers; Scientific computing; Social networks; Stacking; Surveys; Visualization; Web visualization
• ISSN: 2769-5654
• DOI: 10.1109/CSCI58124.2022.00152

Opportunities to conduct advanced surveys and analysis using the Web have been increasing. In such works, an accurate understanding of the human relationships involved with the target of surveys/analyses contributes to their successful execution. However, the scope in which human relationships can be understood is limited due to the rapid expansion of the Web. We have aimed to develop methods for enabling observation of human relationships expressed in descriptions on the Web from various perspectives. In this paper, we describe methods for dynamically extracting human relationships on the Web using closeness centrality. We also describe an experiment to dynamically generate the human-relationship networks in accordance with people who the observer focuses on and discuss the effectiveness of our methods based on its results.