Programmatically computing street intersections using street geometry

摘要

Techniques are disclosed for programmatically deriving street intersections from address data which is presented in textual format, or alternatively, from street geometry data which has been derived from such address data. Preferred embodiments leverage built-in functions and data types of a spatially-enabled object relational database system. Information about streets is analyzed to determine which streets intersect, as well as the intersection locations. This information is stored in a spatially-enabled table for subsequent retrieval. The derived street intersections data supports retrievals which do not rely on proprietary file formats or binary files, thereby enabling faster retrievals and reduced resource consumption requirements. The derived street intersection information may be used in many ways, such as to compute a path from one location to another.