한국고전종합DB 인물관계망 개선 방안

Abstract

In this study, relevant data was collected by crawling information about 67,000 people and more than 200,000 cases of interpersonal relationship information registered in the “Interpersonal Relationship Information” section of the Korean Classics DB. The collected data was verified and processed to develop techniques for visualizing the interpersonal relationship network in 2D or 3D. In addition, in the process of visualizing the interpersonal network, we suggested a direction for improving the human relationship network of the Korean Classics DB, which is currently being serviced. In order to create an interpersonal relationship network that users can access without inconvenience, it is necessary to pre-analyze the quantity and expression method of the data. Since expression in 2D or 3D each has advantages and disadvantages, the amount and distribution of data can greatly affect the direction of visualization. To show a large amount of data on one screen, the person search function and understanding the relationship between two people are also very important development factors. This may vary depending on which method is used, 2D or 3D. Because 3D uses perspective, the search effect is halved by the font size alone. In addition, for people who is called hub data such as Lee Hwang and Song Si-yeol who are interrelated with many people, it is effective to enable users to easily enlarge and reduce. There are many algorithms to search for the relationship between two individuals, but the method of expression must also vary depending on the quantity of data. In addition, the shortest-path searches in computer engineering and shortest-path searches from a humanistic viewpoint may have different meanings. Therefore computer engineering and humanities skill must be combined. In the process of exploring and visualizing the relationship between two peoples, if there is an additional database for government office information or letter data, this can be utilized to more effectively identify the relationship. The basic premise of data visualization such as an interpersonal relationship network is the integrity of the data. During the development process, it was confirmed that there were a large number of errors in the data contained in the interpersonal relationship information in the Korean Classics DB. If the task of correcting data errors, if these errors are also found by a computer algorithm and reviewed by a humanities expert, it would be possible to perform data correction quickly. The interpersonal relationship network developed in this study was created on premise that there should be no inconvenience in use and performance, using the interpersonal relationship information of the Korean Classics DB, and it is the result of the researcher’s use of auxiliary means for big data visualization lectures. It was developed in the form of a web service for an environment similar to the interpersonal relationship information of the Korean Classics DB, which is currently available for service, and the processing time was expressed on screen.