The purpose of this paper is to understand the key factors for efficient maintenance of rapidly aging facilities. Therefore, the safety inspection/diagnosis reports accumulated in the unstructured data were collected and preprocessed. Then, the analysis was performed using a text mining analysis method. The derived vulnerabilities of tunnel facilities can be used as elements of inspections that take into account the characteristics of individual facilities during regular inspections and daily inspections in the short term. In addition, if detailed specification information and other inspection results(safety, durability, and ease of use) are used for analysis, it provides a stepping stone for supporting preemptive maintenance decision-making in the long term.

Abstract
1. 서론
2. 연구 방법론
    2.1 데이터 수집
    2.2 데이터 전처리 과정
    2.3 데이터 분석
3. 분석 결과
    3.1 데이터 전처리 결과
    3.2 터널유형별 데이터 분석
    3.3 지역별 데이터 분석
4. 결론
References

• 서정은(국토안전관리원) | Seo Jeong-eun (Korea Authority of Land & Infrastructure Safety)
• 오진탁(경일대학교 건축학부) | Oh Jintak (School of Architecture, Kyungil University)