NameBase       document scan

To demonstrate, click here to simulate www.csis.org/about/trustees/ in the box.


Enter the address of a web document you want scanned:

         

More on the document scan:
The URL submitted must access a web document, but not PDF, MSWord or anything else that has esoteric formatting. The document should be English-language prose, with names of individuals, corporations, or groups embedded in the text. Lists of names, or names formatted into tables, will probably not work.

Our server fetches the document and extracts the names. The fetch takes up to fifteen seconds, while the extraction takes only a second. The names from the document are linked to possible NameBase matches for each name, if any are found.

Programs that extract proper names from prose are inexact, much like machine language translations. Despite the algorithms, dictionaries, and exception lists, countless possibilities remain that the software fails to anticipate.

home | book reviews | name search | proximity search | country search | register | order