 This paper presents a holistic approach to align heterogeneous geospatial data sets by integrating spatial, lexical, structural and extensional similarity metrics and automatically aggregating them using approval voting. The approach is validated with real geographical semantic webs and reduces the dependency on specific information for more robust alignment under unbalanced distribution of various information. This article was authored by Liu, Pei Yuanqiu, Shiliang Liu and others. Thank you for watching.