112신고내용에 대한텍스트 마이닝 분석과 시사점

Abstract

In order to overcome the limitations of the previous research, text mining analysis was carried out on the contents of 112 calls received for 6 months from a police station of Gyeonggi province. As a result of analyzing calls related to theft, violence, domestic violence crime, keywords related to target, place of occurrence and type of illegal activity were extracted. In the calls of other criminal offenders, business obstruction, drunken driving, voice phishing were detected as important reports and clues were found to set the direction of police activities. In addition, alcohol drinking was extracted as a key word of calls related to disorder, and places with frequent occurrence of disorder and type of illegal behavior were also analyzed. Through text mining, we were able to systematically classify keywords related to the subject, object type, type of action, and place based on the contents of calls. This technique can also be applied to the police box and the clues to the police activities considering regional characteristics were confirmed. In this study, we confirmed the usefulness of text mining in analyzing a large amount of 112 reporting data, but we also confirmed the limitation of in-depth analysis in connection with other reporting related items. Based on the results of the analysis, we suggested 112 report analysis system and crime prediction system that reflects the mining technique, reorganization and expansion of the 112 reporting metrics for applying data mining techniques, strengthening various crime prevention activities through the analysis of 112 report data by crime prevention officer of the police station.