Analysis of the semantic network of post-traumatic stress disorder using Korean social big data

Abstract

Introduction: In this study, we wanted to examine how post-traumatic stress disorder was discussed in Korean newspaper articles with semantic network analysis suitable for unstructured big data analysis.

Methods: This study analyzed 11,304 articles related to post-traumatic stress reported by four major Korean newspapers for three years from July 30, 2017, to July 30, 2020. R 3.6.2 program was used to calculate TF and TF-IDF values, and UCINET 6.0 and interlocked NetDraw was used for DC, EC, and CONCOR values.

Results: As a result of deriving 50 major keywords with high TF-IDF values ​​in newspaper articles related to a post-traumatic stress disorder, TF-IDF values were high in the order of 'sick leave', 'solitary confinement', 'detention center', 'standing order', and 'prisoner'. As a result of conducting a CONCOR analysis to determine which sub-clusters keywords are classified into, the researcher derived each cluster based on words included: 'PTSD by crops' (cluster 1), 'PTSD by broadcasting accidents' (clusters), 'PTSD by farm livestock accidents' (cluster 3), and 'PTSD by various accidents' (cluster 4).

Conclusions Based on the research results, post-traumatic stress disorder needs to be managed nationally. As such, we intend to provide basic data for policy development and intervention programs.