METHOD OF INFERRING SPATIAL MEANING TO TEXT

1 . A computer-implemented method involving a corpus of documents, said method comprising:

analyzing the documents in the corpus of documents to identify known spatial identifiers, each of said known spatial identifiers identifying a corresponding spatial location;

analyzing the corpus of documents to identify a specific text string that is not a known spatial identifier;

analyzing the corpus of documents to assess a degree of association between said specific text string and one or more known spatial identifiers; and

if the degree of association between said specific text string and a known spatial identifier is sufficiently high, associating said specific text string with the spatial location identified by that known spatial identifier.

2 . The method of claim 1 , wherein analyzing the corpus of documents to assess the degree of association involves statistically analyzing correlations.

3 . The method of claim 1 , where the degree of association between the specific text string and a known spatial identifier is measured by determining a frequency with which the specific text string and the known spatial identifier occurs in the same document among the corpus of documents.

4 . The method of claim 1 , wherein said specific text string is a specific word.

5 . The method of claim 1 , wherein said specific text string is a specific phrase.