There are different methods to process duplicate detection in databases, but to process the data quickly at the same time maintaining the quality of database become difficult. In this paper PB and PSNM algorithms are presented to improve the efficiency of duplicate data detection in databases keeping the time to a shorter level.