Title:
LANGUAGE ANALYSIS SYSTEM AND METHOD
Document Type and Number:
Japanese Patent JP3884001
Kind Code:
B2
Abstract:
PROBLEM TO BE SOLVED: To provide a language analysis system which exactly analyzes parts of speech of tokens constituting a sentence.
SOLUTION: A division means 2 divides inputted sentence data into tokens by referring to a dictionary means 8 which preliminarily stores many tokens. A part of speech acquisition means 4 acquires the part of speech for each of the divided tokens by referring to the dictionary means 8. A part of speech selection means 6 determines (or narrows down) the parts of speech of the token into one based on one or more parts of speech before or after (or both) when a plurality of acquired parts of speech exist for each token. In addition, capacity of the dictionary means is drastically reduced and processing speed is increased by a table in which suffixes of verbs and predicate adjectives are commonized.
Inventors:
Shevenko Svetlana
Application Number:
JP2003359781A
Publication Date:
February 21, 2007
Filing Date:
October 20, 2003
Export Citation:
Assignee:
Hiroshi Ishikura
International Classes:
G06F17/27; (IPC1-7): G06F17/27
Attorney, Agent or Firm:
Eio Furuya
Tadashi Matsushita
Hiroaki Majima
Tsurumoto Yoshifumi
Tadashi Matsushita
Hiroaki Majima
Tsurumoto Yoshifumi