Automated Geotechnical Information Extraction from Construction Boring Logs Using Keyword Groups

Abstract

Geotechnical survey data is essential for the construction of civil engineering and architectural structures, with high utilization rates. However, variations in the forms used across different entities such as host organizations, contractors, and structures necessitate manual input tasks during the database creation process, leading to significant consumption of human and time resources. To address this challenge, both standard and distributed boring logs were collected and subjected to comprehensive feature analysis. Based on this analysis, an algorithm capable of automatically extracting the desired geotechnical information from boring logs was developed. This algorithm is versatile, applicable across various formats, and has demonstrated a staggering improvement in information processing speed compared to manual input.