| 摘要 |
PROBLEM TO BE SOLVED: To easily learn and understand knowledge through a knowledge database by processing text information which is transmitted with an input character string together with information included in the knowledge database. SOLUTION: The input character string 12 of a text, a voice, etc., is converted by a semantic preprocessor 17 to a regular input network 18 of meaning units. The semantic preprocessor 17 converts the input character string 12 to the regular input network 18. A semantic processor 19 finds, in the knowledge database, the subset which matches all the semantic units of the input network 18 most. Further, expansion from the knowledge database 11 is performed in relative semantic units. The input network is called a resulting semantic network 13 there. The semantic processor 19 generates a new fractal hierarchic network of semantic units and pointers from the input network 18 and the knowledge data base 11. The resulting semantic network 13 reflects both general meanings of the input character string 12 and individual sides and is represented as the fractal hierarchical network of the semantic units and pointers like the knowledge database 11.
|
