In this study, 27 soil samples were collected for laboratory pretreatment and the total concentration of heavy metal Cd, Cr, Cu, Ni, Pb, Zn, As and Hg was measured. Pearson correlation analysis was carried out on the measured data after removing outliers, and the comparison groups with a significant correlation at the level of 0.01 between the concentration of several groups of elements were obtained. In order to identify effectively source of soil heavy metals by PMF analysis (Positive Matrix Factorization), we drew the location map in the study area and the concentration distribution of heavy metals. Combining Pearson correlation analysis, distribution of heavy metal concentration and PMF analysis, we obtained convincing identification results of heavy metal sources. With C# language and ArcGIS Engine development components, we developed a soil heavy metal database management system to manage the spatial and attribute data needed in source apportionment for soil heavy metals, which will provide data support for the latter sustainable research. In this paper, we proposed a sustainable heavy metal pollution identification research process, SSAPD (sustainable source analysis process based on database), which includes data collection in the field, laboratory measurement, pretreatment, PMF pollution source analysis and database establishment. The process can not only effectively identify the source of soil heavy metal pollution, but also realize the continuity of research and the sharing of data.