[2] To identify place names, GIR systems often rely on natural language processing[3] or other metadata to associate text documents with locations.
[5] After identifying mentions of places and locations in text, a GIR system indexes this information for search and retrieval.
[10] GIR must manage several forms of uncertainty, including semantic ambiguity of mentions of places in natural language text and position precision.
[11] The study of GIR systems has a rich history dating back to the 1970s and possibly earlier.
See Ray Larson’s book Geographic information retrieval and spatial browsing[20] for references to much of the pre-Web literature on GIR.