Duplicate information is often found in soil drilling log databases. In this paper, we propose a method to detect this duplicated information in order to compile a reliable database. The detections are determined by the following method: 1) Each soil name in the soil log is assigned a specific number. 2) This numerized log is then transformed to the spectra of Walsh-Hadamard function. 3) Subsequently, a small number of spectra (approximately 4 or 5) are sorted using a computer spreadsheet program such as Lotus 123 or MS Excel. 4) From the spreadsheets, duplicated drilling logs can be readily identified.