One can also georeference documents by considering similarities in keyword occurrences between a document to be georeferenced and documents that have already been georeferenced and by manual indexing by human experts. The Mammal Networked Information Systems, a network of institutions involved in collaborative georeferencing maintains one of the most extensive normative descriptions of georeferencing methods available [20]. Here we are going to focus on automatic georeferencing based on the contents of the documents text alone.
In an automated approach most projects have based their approaches to georeferencing on a combination of place name identification and natural language processing to identify phrases that modifies the location pointed to by occurrences of place names (“200 km south of the Moskow”) or that provides georeferences that indicates a georeference without actually mentioning a specific place name (“Rosenborgs homefield”).